SID 5 Research Project Final Report
Transcription
SID 5 Research Project Final Report
General enquiries on this form should be made to: Defra, Science Directorate, Management Support and Finance Team, Telephone No. 020 7238 1612 E-mail: [email protected] SID 5 Research Project Final Report Note In line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects. • This form is in Word format and the boxes may be expanded or reduced, as appropriate. ACCESS TO INFORMATION The information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000. Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors. SID 5 (Rev. 3/06) Project identification 1. Defra Project code 2. Project title WU0108 The relationship between water availability and quality in the context of agricultural practices in England and Wales 3. Contractor organisation(s) ADAS UK Ltd 54. Total Defra project costs (agreed fixed price) 5. Project: Page 1 of 29 £ 61,538 start date................. 01 April 2007 end date.................. 31/10/2007 6. It is Defra’s intention to publish this form. Please confirm your agreement to do so. ...................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow. Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer. In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000. (b) If you have answered NO, please explain why the Final report should not be released into public domain Executive Summary 7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work. Changes in the quantity or availability of ground or surface water may arise due to anthroprogenic changes such as abstraction, transfer and impoundment, as well as by natural variability in the rainfall and water balance. Changes in average water quantity, and in the variation within and between years, are predicted to occur over the coming decades due to climate changes. Agriculture places a significant demand on regional water resources, for the irrigation of crops, washing and watering of livestock, and the preparation of produce. These agricultural effects on water quantity or availability, both annually and seasonally, have the potential to impact on water quality, by influencing chemical effects (e.g. reduced dilution of point source inputs of pollutants, or changing the kinetics of in-river processes by changing reach residence times) or by influencing physical effects (e.g. reduced volume of useable habitat for plants and animals). The combined chemical and physical effects can alter the ecology of a river reach, and potentially prevent a reach from achieving the ‘good chemical and ecological status’ required by the Water Framework Directive. The aim of this project was, therefore, to review the relationship between water availability and water quality in the context of agricultural practices, so as to provide Defra (and other stakeholders) with a better understanding and a robust basis for decision making. The specific analyses carried out by this project were: • • • • Summarising the existing standards for good water status that may be impacted by changes in water availability, with reference to the Water Framework Directive; Summarising and mapping data on the agricultural demand for water, and quantifying the agricultural impact on water availability, with special attention to seasonal and inter-annual effects; Providing quantitative evidence of the impact of reduced water availability (due to agriculture) on chemical and ecological status, under present land use and climate; and Investigating the potential impact of climate change on water availability and quality. The results of these analyses were placed in the context of the Water Framework Directive (WFD) and provided an assessment of the regional pattern of impact across England and Wales. A full report is available as Annexe I to this summary report. This project assessed the quantity of water available for abstraction from surface waters for the current SID 5 (Rev. 3/06) Page 2 of 29 (baseline) climate, as well as estimating effects under future climate scenarios. We have been able to place these in the context of the agricultural demand for water. A model was developed for the simulation of time series of flow for the whole of England and Wales. The model is responsive to catchment characteristics of land use and hydrogeology and it showed a good performance over a range of different catchments, without the need for site-specific calibration to observed data. The evaluation of agricultural demands in the context of available water showed that demands from agriculture are significant in the summer period and in certain areas of the country, mainly concentrated in East Anglia. Also, agriculture is presumably the most important abstractor in headwaters, as other abstractors usually abstract from major rivers. In many of the areas in which abstraction of summer surface water is classified as unsustainable by the EA, agricultural demands are not significant. Abstraction of summer surface water is currently considered unsustainable in all of the Southern Region and parts of East Anglia, but in most parts of the Southern Region agricultural demands are less than 3 mm. This is because the crops requiring large amounts of irrigation are not grown to a big extent in this region. In East Anglia, where irrigation demands are significant, summer surface water abstractions are classified as unsustainable in smaller parts, and no additional water is available in any part of East Anglia. With decreasing summer flows under future climate conditions, less water will be available for abstractions, while demands are expected to increase. In terms of water quality, reassessment of standards under the WFD has created closer linkages between chemical, nutrient and hydromporphological status and ecology. Thus, if agricultural water demand impacts on dissolved oxygen, BOD, ammonia or P levels and/or flow, then this could have potential impacts on the ecological status. We have concluded that changes to water quality due to dilution effects are: • chiefly related to point sources (sewage treatment works) rather than diffuse agricultural sources; • are is most likely in the east; • are most likely in the summer months. In terms of climate change impacts on water availability, and the demand for irrigation water in summer, we concluded that: • Climate change is likely to increase demand for irrigation water during summer, and increase the frequency of occurrence of extreme demand years and of low flows; • annual variations are as great as or greater than expected changes in mean value within the next few decades However the change in mean value, and in frequency of demand exceeding availability, will impact on water resource planning and irrigation abstraction licensing. This analysis did not take explicit account of other potential effects of climate change. Greater risk of summer drought, with higher temperatures, could increase irrigation demand: • by extension of Irrigation to a wider range of crops; • by introduction of new crops requiring irrigation; • by extension of irrigation into areas where irrigated crops are currently rare – i.e. towards the west. The extension of irrigated crops towards areas where they currently do not occur could have adverse impact on rivers which are currently of relatively good ecological quality: • by reduction of summer flows; • by quality changes as discussed above; • more directly, by the fact that some of these crops carry with them an increased risk of sediment or other pollutant transfer. The results form a basis of information for input to river basin planning, and for taking account of impacts of climate change, in a uniform way. They should be made available to river basin planning teams. The results suggest that conflict between supply and demand is greatest during summer, and in the east; and that this conflict will become greater rather than less. This suggests that, in future, irrigation may only be viable with winter storage to allow abstraction of water while flows are high. Further assessment of the impacts of climate change should take account of likely changes in land use, and in crops for which irrigation may become economic. As a result of climate change and consequent land management change, and the difficulties of irrigation in eastern areas, it is possible that irrigation will increase in more sensitive western catchments. The risk of this, and the ecological implications, should be investigated. SID 5 (Rev. 3/06) Page 3 of 29 Project Report to Defra 8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer). 1. INTRODUCTION Changes in the quantity or availability of ground or surface water may arise due to anthroprogenic changes such as abstraction, transfer and impoundment, as well as by natural variability in the rainfall and water balance. The water quantity available for use by agriculture is further affected by restrictions placed on abstraction, which are constantly under review. Changes in average water quantity, and in the variation within and between years, are predicted to occur over the coming decades due to climate changes. These could compound low flow problems in summer months especially in southern and eastern regions (Downing et al., 2003). While effects on mean annual water quantity are uncertain and variable, there is consensus that summer rainfall will decrease, causing reduction in summer flows and delay in the start of recharge in autumn. Agriculture places a significant demand on regional water resources, for the irrigation of crops, washing and watering of livestock, and the preparation of produce. The annual average volume of water used for irrigation in 6 3 England is c. 125×10 m (Weatherhead & Danert, 2002) and for the drinking and wash requirements of livestock 6 3 is c. 120×10 m (Smith, pers. comm.). The majority of water for drinking and washing is supplied by the mains and therefore simply adds to the public water supply demand. The dominant source of water for irrigation is abstraction from surface and ground waters, accounting for 95% of irrigation water in 2001 (Weatherhead & -1 Danert, 2002). In the main irrigating regions of Anglia and the Midlands (EA regions), approximately 130 ML day -1 were abstracted from surface waters, and 100 ML day from groundwater sources (Defra, 2005). The majority of the water for irrigation is abstracted during summer (i.e. few users have sufficient storage to carry them through the whole season). Abstractions, therefore, have the potential to significantly decrease natural river flows, and especially low flows in summer months when the demand is greatest. Long-term changes in land use patterns, in response to economic and climate change, may also impact on seasonal river flows by changing the balance of rainfall and crop evapotranspiration through crop type and cover (Hulme et al., 1993). These agricultural effects on water quantity or availability, both annually and seasonally, have the potential to impact on water quality, by: • Influencing chemical effects ~ Including a) reduced dilution of point source inputs of pollutants, and b) changing the kinetics of in-river processes by changing reach residence times; • Influencing physical effects ~ Including a) reduced volume of useable habitat for plants and animals, through changes in flow depth and velocity, and b) changes in water temperatures. The combined chemical and physical effects can alter the ecology of a river reach, and potentially prevent a reach from achieving the ‘good chemical and ecological status’ required by the Water Framework Directive (WFD; 2000/60/EC). The aim of this project was, therefore, to review the relationship between water availability and water quality in the context of agricultural practices, so as to provide Defra (and other stakeholders) with a better understanding and a robust basis for decision making. SID 5 (Rev. 3/06) Page 4 of 29 By ‘water availability’, we have interpreted this to mean water available for use in agricultural production (crop, livestock and amenity plant production systems). However, this always has to be set against the quantity of water within a catchment (surface and groundwater systems) and how agriculture interacts with the catchment water resources. 1.1. Objectives 1. To summarise existing standards for good water status that may be impacted by changes in water availability, with reference to the Water Framework Directive; 2. To summarise and map data on the agricultural demand for water, and to quantify the agricultural impact on water availability, with special attention to seasonal and inter-annual effects; 3. To quantify the impact of reduced water availability (due to agriculture) on chemical and ecological status, under present land use and climate; 4. To investigate the potential impact of climate change on water availability and quality; 5. To use these data to review the regional pattern of impact across England and Wales, with special attention to differences between types of pollutants. 6. To set these finding in the context of the Water Framework Directive and achieving its objectives. 2. APPROACHES The specific analyses carried out by this project were: • Summarising the existing standards for good water status that may be impacted by changes in water availability, with reference to the Water Framework Directive; • Summarising and mapping data on the agricultural demand for water, and quantifying the agricultural impact on water availability, with special attention to seasonal and inter-annual effects; • Providing quantitative evidence of the impact of reduced water availability (due to agriculture) on chemical and ecological status, under present land use and climate; and • Investigating the potential impact of climate change on water availability and quality. The results of these analyses were placed in the context of the Water Framework Directive (WFD; 2000/60/EC) and provided an assessment of the regional pattern of impact across England and Wales. A full report is available as Annexe I to this summary report. 3. RESULTS AND DISCUSSION 3.1. Potential impacts of reduced water availability There are many potential impacts of reduced water quantity in catchments, and these must be taken into account in determining both quantity and timing of any abstractions. In addition to the impacts directly on the river systems and particular reaches affected by abstraction, there can be knock–on effects felt in lakes and other water bodies fed by river suffering low flow events. Reduced flow through lakes and reservoirs can increase the risk of algal blooms. River margin habitats may suffer if river levels or river water quality change. Restrictions on abstractions are often driven by the need to maintain such habitats. Potential impacts include: • Habitat change and river morphology - changes in the natural fluctuations in flow will affect species composition. Greatest concern has focused on low flows in summer, but changes in the pattern of flow regimes could also have important impacts. • Increased temperature and reduced dissolved oxygen - low water levels and low flow rates result in river water temperatures approaching ambient. Under the high temperatures typical of summer, and especially of drought periods, this can result in water temperatures that kill or reduce the viability of fish, and other organisms. At elevated temperatures, levels of dissolved oxygen (and other gases) fall. • Sedimentation - The deposition and accumulation of fine-grained sediment on river channel beds is increasingly recognised as an important environmental problem. Increased sediment deposition overall could occur where the frequency and extent of high flows, with their natural scouring effect, is reduced, especially where this is coupled with lower flows in summer. Depletion of groundwater reserves could have this impact on groundwater-fed streams, causing reduced flow throughout the year. Stagnation during extreme low flows SID 5 (Rev. 3/06) Page 5 of 29 can result in evaporation with total deposition of all suspended matter, leaving a layer of very fine sediment which blocks air access to river beds. • Impacts on estuaries and tidal reaches - Large increases in salinity and other characteristics may occur in near-estuarine reaches during periods of very low flow, when estuarine influence reaches further up the river than normal, with adverse effects on salt intolerant plant and animal species. • Concentration of pollutants - reduction in river flow, coupled with maintained inputs of pollutants, will result in increased concentrations of pollutants. • Sewage Treatment Works - The main issue is that input of sewage-derived pollutants is fairly constant throughout the year, whereas the flow of water (and associated pollutants) derived from the land is generally reduced in summer. Upstream abstractions could limit the quantity of relatively clean water available for dilution of a downstream input of sewage. This risk is greatest in summer, when diffuse inputs and flow from land are at a minimum. • Diffuse agricultural pollution - Diffuse pollution from agriculture occurs mainly during the high flow period of winter. In general, concentrations of pollutants derived from agriculture are less likely to be affected by abstraction of water than are pollutants derived from sewage, since agricultural pollutant inputs are ‘diffuse’ across the landscape, and the abstracted water removes them pro-rata. In livestock farming areas, there can be a problem of intermittent ‘diffuse-point’ source pollution from direct discharge of yard and parlour washings. Such pollution, although more extensive during winter, could occur at any time of year and its impacts will be aggravated by low stream flow volumes. • In-stream transformations - during periods of low flow, residence times are extended. Increased residence times allow greater equilibration with bed sediments, which may allow a decrease in concentration of pollutants such as phosphorus if they exceed that of the bed sediments. Increased residence time also allows opportunities for death and decay of FIOs and decay of agrochemicals in situ. High flows can, however, mobilise sediment and associated phosphorus retained during low flows, through resuspension. • River plant communities - limited dilution or increases in sewage discharges can lead to increases in inorganic nutrient concentrations, in particular P and possibly N, to which phytoplankton, macrophytes and phytobenthos are highly responsive. Nutrient enrichment can lead to an increase in biomass in stream to the exclusion of nutrient-sensitive slow growing species, thus changing the in-river species composition. The initial nutrient concentration in the water column governs the effect of the nutrient enrichment, which may have no impact if the nutrient is not limited. Regarding algal blooms, unless the residence time of the water in the river is greater than the generation time of phytoplankton, which is of the order of several days, significant biomass quantities cannot be achieved. The main risk of algal blooms is therefore in reservoirs, canals and very slow moving rivers (Deflandre & Jarvie, 2004). • Impacts on invertebrates - invertebrates are frequently used as indicators of river health as they exhibit rapid responses to a wide range of environmental factors. The invertebrate community is particularly affected by oxygen concentrations as most are unable to regulate their oxygen uptake. • Impacts on fish - low flows and their corresponding impacts on river velocities and water depths affect different fish species in different ways, the result of which can be changes in species assemblages in a given river reach. For salmon, low flow conditions associated with droughts, coupled with increased water temperature and reductions in dissolved oxygen have serious impacts on fish mortality resulting in gaps in the age structure of communities. Only a few days of conditions that do not meet ideal requirements can result in fish losses. Low flows at critical times can also affect migration; and spawning. • Abstracted water quality – reduction in summer flows could result in elevated concentrations of FIOs in summer. These could pose a danger where the water is used for irrigation of fruit or vegetable crops to be eaten raw and unprocessed. Increases in P concentrations in summer, due to flow reduction, could increase the risk of algal blooms and the extended residence times at low flows would increase this risk still further. Clearly, there are many impacts of water quantity availability on the hydrology, physics, chemistry and ecology of rivers. Abstractions are licensed to ensure that given environmental standards are maintained in rivers defined in terms of discharge rates that are related to other ecological measures. Standards for rivers therefore exist for the UK driven by the EU legislation of the Nitrates Directive, Fisheries Directive, Habitats Directive, Bathing Waters Directive, Shellfish Waters and Shellfish Hygiene Directives. However, the Water Framework Directive (200/60/EC) currently places a greater onus on Member States to examine the links between ecological condition, hydromorphology and flow, and to define a new raft of standards and methods to assess and monitor quality. SID 5 (Rev. 3/06) Page 6 of 29 3.2. Summary of existing standards (Objective 1) The aim was to summarise existing standards for good water status that may be impacted by changes in water availability, with reference to the Water Framework Directive. It is proposed that the current management and assessment methods used for water resources (e.g. CAMS) and water quality (e.g. General Quality Assessment, GQA) in the UK will in time be replaced by a new set of methods to determine whether Good Status is being achieved under the WFD. The new approach will be broader, due to the need to consider the water environment as a whole by considering water quality, quantity (e.g. river flow) and physical habitat (e.g. channel form and depth) alongside ecological indicators (e.g. populations of fish, macroinvertebrates, and macrophytes) and by covering rivers, lakes, estuaries, coastal waters and groundwater. Under the WFD, ecological quality is judged on the basis of the degree to which the observed conditions deviate from those that would be expected in the absence of significant anthropogenic influence (Carvalho et al., 2002) and, as such, ‘good’ and other status classes are assigned to each water body. The current and proposed classification methods and standards for the WFD for both water quality and water quantity are set out below. The focus is on rivers, because abstraction from groundwaters will have little impact on groundwater chemistry, and any ecological impacts will be chiefly impact on surface waters. Specific pollutants are substances (metals, solvents and some pesticides) discharged into the water environment in ‘significant quantities’ and are known to harm ecology. To achieve Good Ecological Status under the WFD, the UK is responsible for developing Environmental Quality Standards (EQSs) for Specific Pollutants. Priority Substances and Priority hazardous substances are of concern in reaching the ‘Good Chemical Status’ classification under the WFD. These substances are especially toxic. The standards for these substances are being set by the European Commission directly (www.wfduk.org). Specific pollutants and priority substances are not considered in any further detail within the scope of this project. Current Water Quality Standards - Environment Agency GQA standards The Environment Agency uses the General Quality Assessment (GQA) Scheme to consistently assess the state of water quality of rivers and canals throughout England and Wales and to assess changes in state over time. More than 7000 sampling sites in England and Wales are assessed for their chemical and biological status, nutrient levels and aesthetic water quality. Chemical, biological, nutrient and aesthetic parameters are monitored across this network (Table 1). The results are compared to predetermined standards and a GQA grade is assigned for that parameter. Table 1. Summary of the GQA. Parameter Determinands GQA scoring system Chemical Dissolved oxygen, Biological oxygen demand & ammonium Score A-F, based on worst of the 3 chemical indicators (A=best) Nutrient Phosphorus and nitrate Score 1-6 for each (1= lowest) Biological Calculation of an Ecological Quality Index based on populations of macroinvertebrates Score A-F (A=best) Aesthetics Litter, odour, colour and presence of oil on surface Score a-f (a=best) Environment Agency Flow Rules The Environment Agency also manages abstractions in England and Wales. Key indicators have been the Q95 index, which is the flow that is equalled or exceeded 95% of the time, or indices of rarer events such as the mean annual minimum flow. The Q95 has often been treated as a hands-off flow, by which no abstraction is permitted when the flow is below this level. Whilst these approaches are not specifically chosen with ecological considerations in mind, there are cases where more generic relationships have been identified to protect specific elements of the river system. Water Framework and Habitats Directives, amongst others, have recently moved the focus towards habitat protection and more complex abstraction and flow management. This is described later. The Catchment Abstraction Management Strategies (CAMS) process (Environment Agency, 2002) was developed by the Environment Agency to ensure that the needs of the abstractor are met, as well as safeguarding the environment. The CAMS process sets environment flow objectives in a consistent way across England and Wales using the default framework of the Resource Assessment Management (RAM) framework coupled with input from catchment stakeholder groups. The CAMS process is repeated every six years. The CAMS is being applied across England and Wales to determine ecological sensitivity of water resources as a forward look to the implementation of the WFD. SID 5 (Rev. 3/06) Page 7 of 29 To assess a river’s sensitivity to a reduction in flow, four elements of the ecosystem are assessed: 1) Physical characterisation; 2) Fisheries; 3) Macrophytes; 4) Macro-invertebrates. Each element is given a score of 1 to 5, with 5 being the most sensitive to change and 1 the least. The scores for each element are then combined to categorise the river into one of 5 Environmental Weighting Bands, where band A is the most sensitive and band E is the least. In turn, the RAM framework specifies allowable abstractions at different points in a flow duration curve for natural flows for each weighting band. The percentage of natural Q95 that can be abstracted for each band is given in Table 2. Table 2. Percentage of natural Q95 that can be abstracted from each band on the RAM framework. Environmental weighting band % of Q95 that can be abstracted A B C D E Others 0-5 5-10 10-15 15-25 25-30 Special Treatment This method focuses on producing an ecologically acceptable flow duration curve and is most appropriate when the river ecosystem is controlled by broad characteristics of dry season/wet season or winter/summer flows (Dyson et al., 2003). A more detailed habitat modelling approach, which aims to directly relate the changes in flow regime to responses of target species, is recommended when more accurate environmental flows need to be defined. PHABSIM is one such method and is outlined below. Importantly, in the WFD context, the CAMS approach does, however, incorporate a ‘ground-truth’ review that considers whether the current condition of the ecology is consistent with that expected from a comparison of actual and allowable abstraction. Proposed Methods and Standards under the WFD In 2004, the UK Technical Advisory Group on the Water Framework Directive (UKTAG) initiated a review designed to lead to standards that would support Good Ecological Status under the WFD. The review covered rivers, lakes, estuaries and coastal waters and both water quality and quantity standards and a framework to support decisions about the morphology of water bodies (UKTAG, 2006). Part of the objective is to develop standards which are readily assessed (e.g. relating to chemistry and flow) to support achievement of the biological objectives. This process will be a continuing one as experience and evidence develop, and a review is envisaged by 2012. The review built upon or augmented existing methods and standards where possible. Many of these will eventually be subsumed within the WFD framework. This UKTAG work has been supported by close working with other EU member states (‘intercalibration’). Surface waters have been grouped of the basis on natural characteristics which are likely to affect their typical ecological communities (altitude, location, geology and size). Environment standards will be type-specific where appropriate. Ecological status The Water Framework Directive has as its default objectives: • prevent deterioration of the status of all surface water and groundwater bodies; • protect, enhance and restore all bodies of surface water and groundwater with the aim of achieving Good Status for surface water and groundwater by 2015. The WFD Annex V, Section 1.2 lists general definitions for rivers, lakes, transitional waters and coastal waters for Good status/potential as follows: “General Conditions”: Temperature, oxygen balance, pH , acid neutralising capacity, transparency and salinity, do not reach levels outside the range established so as to ensure the functioning of the type specific ecosystem and the achievement of the values specified above for the biological quality elements. “Nutrient concentrations”: Do not exceed the levels established so as to ensure the functioning of the ecosystem and the achievement of the values specified above for the biological quality elements. Chemical standards Chemical standards have been reviewed in the light of ecological information on their impact, where available, and scientific expert judgement where not (UKTAG, 2006). The data were derived by identifying, on sites with th Good status, the 90 percentile value of the parameter. Some are as means, others as percentiles. SID 5 (Rev. 3/06) Page 8 of 29 The river typology classifies by altitude (encompassing gradient) and alkalinity (Table 3). The complexity of the classification varies with the parameter. The critical chemical standards identified for rivers are similar to those monitored under the GQA: Dissolved Oxygen, BOD, Ammonia, pH and Phosphorus (orthophosphate) – see th th Tables 4-6. The standard for pH (5 and 95 percentile) should be >=6 to <= 9 for high or good ecological status. Table 3. River Typology for oxygen status and ammonia (UKTAG 2006). Site Altitude Under 80 metres Over 80 metres -1 Alkalinity (as mg L CaCO3) < 10 10 to 50 Type 1 Type 2 50 to 100 100 to 200 Type 3 Type 5 Type 4 Type 6 > 200 Type 7 Table 4. Standards for oxygen status and ammonia in rivers according to ecological status (good or high). Type Upland and low alkalinity (Types 1,2,4,6) Lowland and high alkalinity (Types 3,5,7) DO (% saturation). High Good 80 75 70 60 BOD -1 th (mg L , 90 percentile) High Good 3 4 4 Total ammonia -1 th (mg L , 90 percentile) High Good 0.2 0.3 5 0.3 0.6 The typology for nutrients is slightly different from that for oxygen status and ammonia. The UKTAG consider that evidence in relation to Type 3n (lowland calcareous rivers) may not be reliable, and that rivers classed as good status may actually be impacted. Standards for Type 3n are therefore the same as for Type 4n. Table 5. River Typology for nutrients (nitrate, P) in rivers (UKTAG 2006). Site Altitude Under 80 metres Over 80 metres -1 Annual mean alkalinity (as mg L CaCO3) < 50 >50 Type 1n Type 3n Type 2n Type 4n -1 Table 6. Standards for phosphorus in rivers (UKTAG 2006) (soluble reactive P, ug L , annual mean). Type 1n 2n 3n, 4n High 30 20 50 Good 50 40 120 Standards for P were set in relation to diatoms. The revised standards cause a slight increase in failures associated with ammonia, and a slight decrease in failures related to BOD, dissolved oxygen and P. Nitrates standards are currently set by the Nitrates Directive (Defra, 2007) and have not been changed for rivers. Water quantity standards The Water Framework Directive requires hydromorphology, including flow and water level, to be protected for its own sake at sites of High Status. Elsewhere, the criterion is determined in relation to ecological objectives. Criteria will be assessed separately for Heavily Modified Water Bodies (e.g. hydropower schemes), where the objective is to achieve good ecological potential but it is recognised that there are limits to what can be achieved at reasonable cost. An expert consensus approach then defined the proposed regulatory standards and thresholds for flow alteration that would ensure Good Ecological Status (GES) in water bodies, for each river water body type as defined in the new typology. Since all parts of the flow regime have been recognised as having some ecological importance, there has been a move away from low flow measures towards environmental flows in relation to the natural flow characteristics, and various attempts to define environmental river flow requirements have been reviewed by Acreman & Dunbar (2004). The river typology developed for flow conditions (Table 7) differs from that currently used under the Environment Agency’s Resource Assessment and Management Framework (RAM) as a component of the CAMS process. However the standards developed are broadly in line with current criteria. Much will still rest on interpretation in the light of local conditions and competing demands on the water resource. SID 5 (Rev. 3/06) Page 9 of 29 Table 7. Typology for water resources standards for rivers (UKTAG 2006). Type A B C D Gradient (m per km) Altitude (m) Description A1 0.8+/- 0.4 36 +/- 25 A2* Slightly steeper 1.7 +/- 0.8 low altitude 55 +/38 Predominantly clay. South East England, East Anglia and Cheshire plain Chalk catchments; predominantly gravel beds; base-rich Hard limestone and sandstone; lowmedium altitude; lowmedium slope; typically mesotrophic with gravel-boulder or pebble-cobble) bed B1 4.1 +/-9.9 93 +/- 69 B2 Shallower than B1 2.7 +/- 10.7 71 +/- 58 Non-calcareous shales, hard limestone and sandstone; medium altitude; medium slope; oligomeso-trophic with pebble, cobble and/or boulder bed C1 5.4 +/- 6.5 101 +/-84 C2 Steeper than C1 7.3 +/- 10.8 130 +/- 90 Granites and other hard rocks; low and high altitudes; gentle to steep slopes; ultraoligo Oligo-trophic, with cobble, boulder, bedrock, and/or pebble bed D1 Medium gradient 11.3+/- 15.6 low altitude 93+/- 92 D2 High gradient 25.5 +/- 33 High Altitude 178 ± 131 Clay and/or Chalk; low altitude; low slope Eutrophic; silt-gravel bed Hard sandstone, Calcareous shales; Predominantly South and South West England and South West Wales Predominantly North West and East Scotland Hard limestone; more silt and sand than C2; mesotrophic Non-calcareous shales; pebblebedrock; Oligomeso-trophic Oligotrophic, substrate finer than D2 (including silt and sand); more slow flow areas than D2 Stream order 1 and 2 bed rock and boulder; ultra-oligo trophic torrential * To reflect the different sensitivities of the headwaters of chalk streams to the downstream reaches, type A2 was split into two – A2 (headwaters) and A2 (downstream) These river types are defined by the physical characteristics of catchments that control the magnitude of average and low flows. In order to define the river types for England and Wales for this project, the physical characteristics can be derived from summary statistics held in the MAGPIE DSS database. This typology, presented in Figure 1, provided the basis of our investigations of the impact of river low flows presented later in this report. Figure 1. Rivers in England and Wales classified using the UKTAG Typology. SID 5 (Rev. 3/06) Page 10 of 29 The standards for high flow status relate directly to those set by the WFD. Thus for high ecological status, at low flows occurring 5% of the time over a 10-year period, the permitted abstraction is set as a maximum of 5% of natural flow, and even at normal to high flows no more than 10% may be abstracted. However, for good status or below the standards are set in terms of ecological criteria. The standards for good status are more complex. Class A2 is further divided into upstream and downstream classes. The flow regime is partitioned into percentiles, and the permitted percentage perturbation of flow (by assumption, normally a flow reduction) is greater at average to high flows, than at low flows. The standards are generally set to be tighter during summer than winter. They are designed in the light of the need to protect macrophytes in spring and early summer, and macro-invertebrates and fish in late summer to early autumn. ‘Hands off’ flows are expected to continue to be operated by the Environment Agency as required, and the rules may be over-ridden during special conditions such as drought orders. Where flows are enhanced, for example by sewage discharge or pumped transfers between rivers, the resulting flow prior to abstraction is considered the ‘natural’ flow. It will be seen that during low flow periods in summer, flow reductions envisaged in lowland rivers are typically 7.5 to 15% of natural flow (Table 8). However, these restrictions are not applied on a daily basis – it would be possible in principle to take a large amount of water, exceeding these limits, on some days and less or none on others. Additional restrictions may therefore be needed to protect sensitive ecosystems at critical times. Standards for Moderate and Poor status have not been set, but it is envisaged that they may involve an additional abstraction of about 15% of natural flow. Table 8. Water resources standards for rivers and Good Status (UKTAG, 2006). Types A1 A2 (downstream), B1, B2, C1, D1 A2 (headwaters), C2, D2 Salmonid spawning and nursery areas (not Chalk rivers) Season Flow > QN60 Flow > Flow > QN70 QN95 (% change allowed from the natural flow) Flow < QN95 April –Oct 30 25 20 15 Nov –March 35 30 25 20 April –Oct 25 20 15 10 Nov –March 30 25 20 15 April –Oct 20 15 10 7.5 Nov –March 25 20 15 10 April –Oct 25 20 15 10 Nov –March 20 15 flow > QN80 10 Flow < QN80 7.5 Detailed approaches to setting flow regimes The standards under development by UKTAG provide a framework, but more detailed assessment will be required for many rivers. Possible more detailed approaches include: • Dundee Hydrological Regime Assessment Method (DHRAM). • Physical Habitat Simulation (PHABSIM) • Lotic Invertebrate Index for Flow Evaluation (LIFE) More detail on these is presented in Annexe I. Conclusions The introduction of the WFD has placed more of a focus on managing water for maintaining/improving ecological condition of the water. UKTAG have been developing standards that can be used to measure ecological impact and these standards will continue to develop and be used in the implementation of the WFD. For water quality, it is perhaps not surprisingly that the list of chemical and nutrient parameters is similar to that monitored under the GQA: • Chemical: dissolved oxygen, BOD, pH, ammonia • Nutrient: phosphorus SID 5 (Rev. 3/06) Page 11 of 29 Other EU Directives (to be subsumed into the WFD) also provide various standards, most notably the: • Bathing Water Directive – FIOs • Nitrates Directive – nitrate • Bathing Waters Directive - as well as DO, BOD, pH and ammonia, includes zinc and chlorine • Shellfish Waters Directive – pH, temperature, colour, suspended solids, salinity, DO, hydrocarbons, metals, organohalogens, FIOs and saxitoxin For water quantity, UKTAG have developed a framework for permitted flow reductions. For rivers of high ecological status the standards are fairly inflexible. For good status, the framework is more complex, but decisions are based on river typology, season and flow rate. Basically, during low flow periods in summer, flow reductions envisaged in lowland rivers are typically 7.5 to 15% of natural flow. However, these restrictions are not applied on a daily basis – it would be possible in principle to take a large amount of water, exceeding these limits, on some days and less or none on others. Additional restrictions may therefore be needed to protect sensitive ecosystems at critical times. Although this framework provides useful guidance, it may be necessary to undertake a more detailed analysis at the local level. Thus, in summary, although much of the emphasis on water standards has moved to ecological drivers, most of the parameters required in practice have already been monitored under the GQA. The WFD has driven the development of a more stringent framework for linking ecology both to chemical/nutrient parameters and hydromorphology. This provides a more robust framework for interpretation. Consequently, our report has focused on: • Understanding the links between flow and changes in chemical/nutrient quality • The potential impacts of agricultural abstraction on flow, particularly in the summer 3.3. Agricultural water demand (Objective 2) The aim was to summarise and map data on the agricultural demand for water, and to quantify the agricultural impact on water availability, with special attention to seasonal and inter-annual variation. Agricultural water demand Agriculture places a significant demand on regional water resources, for the irrigation of crops, washing and watering of livestock, and the preparation of produce. The annual average volume of water used for irrigation in 6 3 England is c. 125×10 m (Weatherhead & Danert, 2002) and for the drinking and wash requirements of livestock 6 3 is c. 156×10 m (King et al., 2006). The majority of abstractions from surface and ground waters in England and Wales are for the public water supply (45%) and electricity supply (30%), with agriculture accounting for only c. 1% of total actual abstractions (EA, 2005). The dominant source of irrigation water is abstraction, with surface waters accounting for 58%, and groundwater accounting for 36% of irrigation water in 2001 (Weatherhead & Danert, 2002). Water for drinking and wash requirements of livestock are normally obtained from the mains supply. In the main irrigating regions of -1 Anglia and the Midlands (EA regions), approximately 130 ML day are abstracted from surface waters, and 100 -1 ML day from groundwater sources (Defra, 2005). The usage of water for irrigation is obviously concentrated in the drier parts of the country and peaks during the summer months, and so demand is usually greatest at those times and in those regions where water availability is low. The overall impact of agricultural water demand may therefore be significantly greater than the contribution to the national abstraction budget implies. Crop irrigation Irrigation water is assumed, in water resource calculations, to be entirely consumed. In contrast, water taken for many other purposes (washing, household uses) is largely returned to surface waters, via water recovery/sewage treatment works. The irrigation demand occurs almost entirely during summer (apart from a small usage for frost protection of fruit). It therefore places pressure on water supplies when rivers are at their lowest. The relative importance of irrigation as a consumer of water is increased by this. Temporal elasticity can be improved by on-farm water storage. Although 40% of root and fruit farmers who use irrigation have reservoirs, 80% of those using water abstract only during summer, indicating that most of the storage capacity is insufficient to cover the whole summer; 11% abstract only during winter, and the remainder abstract throughout the year. Weatherhead et al. (1997) reported an increase in the total number of on-farm reservoirs and a doubling in the total storage capacity in the 5 years from 1990-95. However, 90% of the water used in 1995 still came from summer abstraction. In the 2005 survey, 42% of holdings now have reservoir storage capacity, and 30% of the water used came from stored sources (50% of their capacity). SID 5 (Rev. 3/06) Page 12 of 29 Irrigation demand obviously varies with the weather. It is both highly seasonal (concentrated in the summer) and more variable from year to year than other agricultural water uses. The financial justification for installing an irrigation system rests in the drier years when yields of un-irrigated crops are low, and prices high. Economics dictate that sufficient equipment is installed to deal with these dry periods. During such periods, applications can exceed 25 mm per week on each irrigated field and the equipment will be in daily use, moving round the irrigated area on the farm in a 3-5 day cycle. During such periods, any interruption in water supply risks loss of yield on part of the farm. The reported actual abstraction data show that average abstraction for irrigation is only 39% of the total licensed volume (Defra, 2000). The license needs to be issued to cover the design driest year. Of the irrigated field crops, more than half of abstracted water is used for potatoes, and the next biggest use of irrigation is for field vegetables (Weatherhead & Danert, 2002). These crops are both high value, and yield and quality are highly responsive to irrigation water (Bailey, 1990). At present, irrigation installation is most commonly economically based on these crops, and other crops are irrigated opportunistically in years and at times when the potato and vegetable crops on the farm do not require irrigation – for example, early and late in the season. The national irrigation area is increasing slowly, at about 0.6 to 0.8% per annum based on current survey trends. The volume of water used is predicted to increase more steeply, with greater emphasis on the most profitable crops and greater average volumes applied. The national predicted growth rate in actual volumetric demand for a ‘design’ dry year was estimated to be about 1.5% from 2001 to 2021 (Weatherhead et al., 1997). In addition to this trend, which was calculated after correction for weather patterns, climate change must be taken into account. Warmer and drier summers will tend to increase irrigation demand on those crops currently irrigated. Livestock requirements The water requirements for livestock drinking water and washing were reviewed by Thompson et al. (2006). The 3 total estimated annual water use for livestock is 156 million m . The cattle sector is the major consumer, with a 3 3 3 total requirement of c. 82 million m , followed by sheep, at 17 million m , poultry at 12 million m and pigs at c. 8 3 million m for England and Wales. The predominant use of water is for drinking, and usage is concentrated in, but not confined to, the wetter parts of the country where most cattle are raised. Expressed as a percentage of total water, drinking water requirements ranged from 79% for dairy cattle, 87-99% for different categories of pigs, >99% for sheep and 96-99% for poultry, respectively. The main source of water for livestock is mains water, although 21% of water used for dairy stock (7% of total used for livestock) is from other sources (King et al., 2006). The contribution of livestock water demand is therefore only 8% of total direct agricultural abstractions from surface waters nationally. National Context The reported quantity of water distributed via the water companies of England and Wales averaged 15,300 ML -1 6 3 day during 2001-2006, or about 5600 ×10 m per year (OFWAT, 2006). Agriculture accounted for about 2.6% of this distributed water, mainly for livestock use. Direct abstractions for irrigation and other agricultural uses accounted for only c. 1% of total licensed abstraction volume, and 0.6% of total actual abstraction in 2004. They represented a volume equivalent to 2% of the abstraction for public water supply. However, irrigation water places greater pressure on water supplies than indicated by the annual volumes used because irrigation requirement is often maximal at periods of minimal supply, i.e. is temporally inelastic; and is normally abstracted directly from river or groundwater, making it spatially inflexible. Irrigation demand is at present localised in space and time. During hot dry weather, irrigation demand may exceed 25 mm per week on irrigated land. During summer, river flows are commonly as low as 10% to 20% of the annual average. For a river with an annual runoff of 200 mm (typical of much of arable England), a flow of 10% of the mean is equivalent to 0.4 mm per week. It can readily be seen that irrigation of even 1% of the catchment land area would, at periods of low flow, use more than half the flow – well in excess of allowable abstraction. In contrast, public water supplies are drawn from multiple sources, and have systems of storage and transfer to help maintain supplies during drought conditions. Mapping Agricultural Water Demand The general approach to mapping agricultural water demand was to integrate existing coefficients of water use per unit crop area and livestock place (Thompson et al., 2006; Weatherhead & Danert, 2002) with agricultural census data for the year 2000 held in MAGPIE (a Geographic Information System; Lord & Anthony, 2000). Water 2 demand was calculated at a spatial resolution of 1 by 1 km and summarised for each of the Water Framework Directive river catchments defined by the Environment Agency. Irrigation water demand was flat profiled during the summer months (June to August) based on actual abstraction records (King et al., 2006), whilst water demand for livestock requirements was spread evenly throughout the year. It is likely that a flat profile of water demand would be observed for the pig and poultry sector, with production dominated by mostly year-round SID 5 (Rev. 3/06) Page 13 of 29 housing systems, but for sheep and cattle peak demands for drinking water are expected during the summer months. Irrigation Demand Coefficients - Crop water use coefficients by crop category and region were calculated using the irrigation survey and census data. The Irrigation Survey 2001 (Weatherhead & Danert, 2002) provides the irrigation usage for field crops (i.e. not protected crops and nursery crops). The data collected is the area of irrigated crops and amount used for various crop categories. Thereby, an average volume of water used per unit area of irrigated crop can be calculated. The volume of water used per unit area of all cropping was then estimated by combining this data with the total crop areas reported by the agricultural census (Data shown in Annexe I: Table 4.2). As irrigation water demand is dependent on summer weather, the survey data always also reflect the weather conditions of that particular year; for example in 2001 irrigation water demand was comparatively small. Using a methodology developed by Weatherhead et al. (1994), Downing et al. (2003) estimated the potential irrigation water use if 2001 had been a ‘design dry year’, defined as a dry year with an irrigation need exceeded in 1 out of 5 years. This information was used to calculate crop water use coefficients for the irrigation demand in a design dry year (Data shown in Annexe I: Table 4.3). Livestock requirement coefficient - For livestock consumption, estimates of the typical water use per head of livestock were taken from the recent Defra report “Water Use in Agriculture: Establishing a baseline” (King et al., 2006). This report used data available from recent studies and surveys, as well as guidelines on “typical stock requirements”, which are available as part of “Best Practice”, published by the Environment Agency. These data were summarised to give coefficients of drinking and wash water requirement per head of livestock according to the Defra livestock census categories (Data shown in Annexe I: Table 4.5). Some animal categories had to be considered together as not all categories identified within the Defra census fit neatly within the groups readily recognised within commercial production for which data on water requirements had been derived. Thus, beef store cattle and growers as dairy replacements were considered together. Similarly maiden gilts and “barren sows fattened for slaughter” are growing pigs and were considered together with finishing pigs in terms of drinking water, and with dry sows for wash water since they are likely to be relatively few in number and kept on straw rather than a slurry system. Figures 2-4 map the total annual water use of livestock and irrigation, expressed as mm water over the whole contributing catchment. The irrigation demand is based on records of actual water use in 2001, whilst the livestock water use is an annual average derived from usage coefficients. Water demand is averaged over the total contributing catchment area (from the mouth of the catchment to the headwater), but mapped against the most downstream subcatchment. The demand that is shown at the mouth of a catchment is actually the averaged demand over the total contributing catchment area. This contributing area may be as great as 1000 km² for major rivers. This approach is used for all maps in this report. It is unsurprising that the highest values for agricultural water demand are located within East Anglia, where water is mainly used for crop irrigation (Fig. 3). The maximum annual demand is equivalent to 5-10% of the annual average runoff from the river catchments, and is expected to contribute significantly to the demand on available resources in this region. The demand for livestock is greatest in the intensive grassland areas of Wales and the west of England, but the overall demand is generally less than 1% of the annual average catchment runoff (Fig. 2). Overall water demand for irrigation is increased by c. 40% in a design dry year (Fig. 4). Figure 2. Map of calculated average annual agricultural water use for livestock, expressed as a depth of water (mm) over the total river catchment area. SID 5 (Rev. 3/06) Figure 3. Map of calculated average annual agricultural water use for irrigation in 2001, expressed as a depth of water (mm) over the total river catchment area. Page 14 of 29 Figure 4. Map of calculated average annual agricultural water use for irrigation in a design ‘dry year’, expressed as a depth of water (mm) over the total river catchment area. Conclusions The aim was to summarise and map data on the agricultural demand for water, and to quantify the agricultural impact on water availability, with special attention to seasonal and inter-annual variation. The reported total water usage by agriculture (for livestock and for irrigation) is quite small – approximately 1% of total abstractions, with the following breakdown (on average): Irrigation of crops Watering/washing of livestock Glasshouse/nursery stock 3 125 million m 3 156 million m 3 <20 million m However, when considering the effects of agricultural water demand on water resources, there are several factors to consider: • • • Seasonality – whereas, the watering requirements of livestock production will be reasonably uniform through the year, irrigation requirements are highly seasonal with (obviously) most applied in the summer. This is at a period when river flows will also be at their lowest, thus maximising any impacts of water abstraction. There is no scope for shifting time of applications. Irrigation must be applied at critical growth stages and soil moisture deficits, which generally fall in the May-August period. Thus, although livestock requirements are greater in total, the highly seasonal nature of irrigation means that the potential impact is much greater. Spatial variation – by combining agricultural statistics with livestock and crop irrigation coefficients, we have been able to identify areas of the country with the greatest water demand. Not surprisingly, this is in the east for irrigation and the west for livestock production. This spatial distribution, like the temporal, compounds the potential impacts of irrigation abstraction. The livestock water requirements generally occur in the wetter regions and so only constitute a water requirement of about 1% of a catchment’s total water run-off. In contrast, irrigation requirements in the east can require 5-10% of a catchment’s total run-off (most of it in summer when flows are at their lowest). Inter-annual effects – Environment Agency data show that the livestock water requirements are generally similar between years, whereas irrigation requirements can vary considerably, depending on summer weather. Annual demand can vary by a factor of two; and, of course, demand is much greater in a dry year when river flows will also be low. Our spatial analysis showed that irrigation requirements in a design dry year increased irrigation demand by 40%, increasing the pressure on water resources in the east, but also in the main irrigation areas of the midlands. Thus, in summary, the irrigation requirements for high value crops (particularly in the drier east) has the potential to greatly impact on the water resources of these catchments and, in particular, to impact on low flows. Further analysis to quantify these effects was undertaken when considering the impacts of climate change. The model constructed for this purpose also provided baseline (current day) calculations based on a long-term weather dataset. The resulting modelled baseline flows and ‘available water’ for surface waters across England and Wales were compared with agricultural demand. Available water was defined as the water that was available for abstraction after taking account of permitted levels of abstraction from surface water for complying with good SID 5 (Rev. 3/06) Page 15 of 29 ecological status as defined by UKTAG (2006). The quantity of water currently abstracted for other purposes such as public water supply was not taken into account. That is, all users would be in competition for this ‘available water’. Simulations based on the long-term weather dataset clearly demonstrated the spatial variability across England and Wales with flows of 60 mm in East Anglia ranging up to 3000 mm in parts of Wales and north east England – with variation (unsurprisingly) matching rainfall distribution. In terms of available water, the average across England and Wales was 100 mm for the year, again with large variability (<30 mm in 20% of East Anglia). However, it is the availability of summer flow (June- August) that is of great importance when considering irrigation requirements, as described above. The average available water was 7 mm (range 1-100 mm). The simulations confirmed also that, based on annual totals of available water, there is no conflict between available water and agricultural demand. However, for the summer months from June to August, there are a number of catchments, where the agricultural demand is greater than the summer available water. These catchments are mainly located in East Anglia, where the summer demand for irrigation exceeds the summer available water in c. 30% of the area. This area is extended in a year with a dry summer (defined as a year with a summer flow which is exceeded in 80% of the years), when the irrigation demand exceeds the available water in c. 60% of East Anglia. In the areas where the demand is in excess of the available water, the demand averages 260% of the available water in an average year and 440% in a dry year. These conclusions are necessarily preliminary. Two assumptions which would lead to an over-estimation of the problem are: • • It was assumed that all of the water was abstracted from surface waters, whereas it is only 60% that comes from this source. It was assumed that all water was abstracted in the summer months for irrigation, whereas about 10% is supplied from on-farm reservoirs (and this proportion is increasing) that are filled during winter. There was some indication from the test catchments that the model may slightly underestimate flow. Conversely, the monthly time step of the calculation is equivalent to assuming some degree of storage and flexibility in timing of abstraction, which is not currently the case on all farms. Therefore, there will be years when total summer flow is sufficient to meet demand, despite inadequate flow having occurred during part of the summer. In addition, conflicts between summer flow and demand are likely to become more acute, due to a combination of climate change; and tightening of ecological standards for river flow and water quality. Despite these caveats, the approach highlights the challenges in some parts of the country in supplying sufficient water to meet agricultural requirements at key times of the year without disrupting flow regimes with potential impacts on the ecological status of some waters. It should also be considered that there are other abstraction pressures on waters and agriculture is a relatively small component. Abstraction for public water supply and industry is far greater in volume than for agriculture; but in the long-term their relative impact is moderated by the fact that most of the water is recycled and returned to the system. 3.4. Impacts of reduced water availability (Objective 3) The aim was to quantify the impact of reduced water availability (due to agriculture) on chemical and ecological status, under present land use and climate. Ground and surface water abstractions reduce river flows. Reduced flow is associated with reduced dilution of point source inputs of pollutants (mainly sewage, but including industrial and urban discharges). The key indicators of this form of pollution are increased concentrations of P and BOD at low flows. A given volume of abstraction will have a greater proportional impact at summer low flows. Irrigation abstractions from rivers will be concentrated within the summer period, as described earlier. Irrigation abstractions from groundwater will also be predominantly during summer, but will have a more general impact via reduction in the groundwater levels, which will affect flows throughout the year. Water for other agricultural uses including livestock is required all year round, and is largely taken from the mains. It is, therefore, a more general load on the total water supply of an area, and may not be attributable directly to a specific reach or catchment. The objective of this analysis was to present empirical evidence of the relationship between river low flows as affected by abstraction and water quality. The direct impact of agricultural abstractions on river water quality by the reduced dilution of pollutant loadings was expected to be difficult to detect empirically as abstractions are limited to c. 20% of the natural flow. The natural variability in water quality sample results due to variations in timing with respect to flow conditions and the size of the pollutant load was expected to be greater than the magnitude of the dilution effect. We therefore conducted an analysis of the general impact of changes in river flow per se on water quality, from which we could extrapolate the impact of agricultural abstractions. The analysis SID 5 (Rev. 3/06) Page 16 of 29 made use of data from the Environment Agency General Quality Assessment monitoring network. Analyses were made of regional and annual variations in water quality data and river flow based on data summarised for chemical and phosphate status reports. A more detailed analysis of daily variations in quality in relation to sitespecific flow measurements was possible for the smaller number of sites in the Harmonised Monitoring Scheme, which represents generally larger rivers discharging to the sea, with a relatively high point source contribution. Regional relationships between annual flow and water quality Strong negative relationships (i.e. poorer quality at lower flows) were established between river flow and 2 2 chemistry classification for the East of England (r 80%) and London (r 48%) regions and a weak negative 2 relationship for the South East (r 25%). These relationships existed for both annual total flows and the summer flows. These relationships were observed for regions with relatively low river flows and known high effluent discharges due to a concentration of people and industry. In wetter regions, the relationship was more variable with no 2 2 relationship in the South West region (r 7%), and a strong negative relationship in the North West (r 44%) that was possibly due to improvements in effluent treatment rather than variations in flow. In Wales, there was an 2 unexpected positive relationship (r 36%), i.e. poorer quality at higher flows. This might be associated with the increased importance of diffuse agricultural pollution during high runoff years in this more rural region. These analyses confirmed the general relationship between flow quantity and water quality. The results are consistent with the concept that, in areas where water quality is dominated by sewage effluent inputs, dilution is an important determinant of quality, and low flows are associated with poorer quality. Seasonality of water quantity and quality River monitoring sites were classified by region, and average values of the determinants of interest and normalised flow were calculated for each calendar month. Regional and national averages of flow and water chemistry were calculated from site mean data, in order to avoid giving undue weighting to sites with more measurements within the periods of interest. Lowest flows occurred on average during July and August, These are the prime months for irrigation. Lowest flows were later, during August /September, in Southern Region and Thames. Both these regions have a high proportion of groundwater-fed rivers. Groundwater contributions will maintain flows in early summer, declining during later summer. Lowest mean monthly flows were about 40% of average overall, with relatively little variation between regions despite the wide range of climate and other factors. On average, greatest concentrations of orthophosphate (OP) occur in July to September (Fig. 5). These are the months of lowest flow. The concentrations during the summer were on average twice the smallest concentrations, which occurred during January. This broad pattern, that OP concentrations are greatest when flows are smallest, is consistent across all regions. In Southern and Thames region, the maximum concentrations occur somewhat later (September) compared to other regions, which is consistent with the later minimum in flows. These patterns are characteristic of waters where the phosphorus concentrations are dominated by sewage inputs (Jarvie et al., 2006). Mean OP concentrations are greatest in the Midlands, Anglian and Thames regions, which are areas of relatively low flow and high population density. They are least in the South West and Wales. However, even in these regions, the HMS data show a tendency for increased concentrations of OP during summer. Many of the HMS were deliberately sited downstream of sewage works, and the majority are on major rivers. The incidence of sewage contamination is therefore relatively high. 1 1.5 0.75 1 0.5 0.5 OP, mg/l P Normalised flow 2 Normalised flow OP, mg/l P 0.25 0 0 0 3 6 9 12 Month Figure 5. Seasonality of flow and OP concentration: England and Wales HMS sites. SID 5 (Rev. 3/06) Page 17 of 29 5.2.2 Variation between years The analysis was repeated to examine the year-to-year variation in flows and concentrations overall. Variation in (normalised) flow during July to August was significantly correlated with July-August mean orthophosphate 2 concentrations (r = 0.37), as shown in Figure 6. Orthophosphate 1.40 R2 = 0.37 1.20 1.00 0.80 0.60 0.40 0.20 0.00 0.00 0.20 0.40 0.60 0.80 1.00 Me a n Norma lise d Flow Figure 6. Relationship between mean July-August OP concentrations in different years and July-August normalised mean daily flow. In addition to OP, concentrations of the following during July and August were also negatively correlated with July-August mean daily flow: 2 • chloride (r = 0.39) 2 • conductivity (r = 0.50) and 2 • boron (r = 0.21). All of these relationships indicate dilution behaviour typical of pollutants derived from sewage. For other sewagerelated pollutants, the relationships were less obvious, but there were trends to greater values at low flow and during summer for both nitrite and copper. BOD and oxygen saturation behaved inconsistently in relation to flow. Detailed analysis of data for individual sites indicated a greater risk of very high values at low flows, but the range of values was also wider so that mean concentrations did not behave consistently. The variable patterns for BOD may also reflect the fact that it is also contributed by diffuse sources especially at high flows. It should perhaps be noted that summers with low flow tended to have greater temperatures also. In contrast to OP and other sewage-related pollutants, concentrations of pollutants derived mainly from diffuse sources did not increase during summer, and were not greater in years of low summer flows. In general, these pollutants showed reduced concentrations during summer and at low flows: • suspended sediment and turbidity; • nitrate Predictive relationships between OP and flow In order to obtain a generalised predictive relationship between OP concentrations and flow during the period when flows are most likely to be low, all HMS sample sites with values for both OP and flow during June to September inclusive were combined. After initial exploratory analysis, detailed regressions were confined to sites with more than 100 measurements for this summer period. The analysis above indicates that OP concentrations should be proportional to the inverse of the flow, and approximately proportional to the quantity of OP added by point sources, where these are important. It has been shown earlier that sites with elevated OP concentrations generally are dominated by point sources, and show an increase in concentration at low flow. The theoretical analysis indicates that, where point sources dominate OP concentrations, concentrations will be proportional to the inverse of flow rate. At moderate to low flows, where diffuse sources become relatively insignificant, both the concentration at any given flow rate, and the rate of change of concentration with inverse of normalised flow rate, will be linearly related to the OP contributed by point sources. For this reason, both flows and OP concentrations were converted to normalised values, by expressing them relative to the temporal mean flow or concentration for all data from that site. Normalised flows between 0.1 and 1 were converted to their inverse, and assigned a flow class in the range 1 to 10 at intervals of 1 unit. For each SID 5 (Rev. 3/06) Page 18 of 29 site, the regression of normalised OP on the inverse of the normalised flow was calculated. The gradient of this regression was found to be linearly related to mean OP concentration, as would be expected from the theoretical analysis. For sites where the regression explained more than 25% of the variance in concentration, the relationship between the coefficients of the regression and mean OP concentration was investigated further. A relationship was derived by expressing all OP concentrations relative to the site temporal mean OP concentration, and averaging by flow class: Cr = (Mean OP) * ( 0.72 + 0.16 / Fr) 2 r = 0.95 Where: -1 Measured total instantaneous OP concentration in the river = Cr mg L P Measured total instantaneous flow in the river = Fr normalised flow units This relationship did not hold for high flows (values of Fr greater than 1), since it predicts OP concentrations somewhat greater than expected from diffuse sources at infinite flow. This is partly because the assumed linearity ignores the dynamics of the contribution from diffuse sources. However, it allows us to predict the relationships to be expected at low flows. As a check on the above relationship, the relationship for the mean of all sites with more than 100 samples taken during June – September was investigated. Samples were classified by flow class, to make comparison between sites simpler. Sites where the correlation coefficient between flow and concentration was smaller than 0.5 were omitted. For the remaining 81 sites, those with few measurements at flows less than 20% of mean flow were classified as ‘groundwater fed’ (28 sites), and only data above 20% flow were used, since the few samples at low flow were too variable to give meaningful trends. For other sites all data above 10% flow were used. There was a consistent difference in behaviour between these two groups of sites, but further sub classification did not yield further differentiation. The resulting relationships were: Groundwater-dominated sites: Cr = (Mean OP) * (0.74 + 0.24/Fr) All other sites: Cr = (Mean OP) * (0.89 + 0.12 / Fr) Within each group, the coefficient of variation of normalised OP at a given normalised flow averaged 31%, -1 compared with 145% for the OP concentrations expressed as measured (mg L P). In other words, classification in this way reduced variability compared to the very wide range of actual OP concentrations. The regressions obtained from these averaged data were very similar to those obtained by averaging the coefficients from individual sites (Fig. 7): 2 Groundwater-dominated sites: Cr = (Mean OP) * (0.79 + 0.20/Fr) r = 0.96 All other sites: Cr = (Mean OP) * (0.71 + 0.14 / Fr) r = 0.98 2 OP / (Mean OP) 2.5 2.0 1.5 1.0 Groundwater-fed 0.5 Other 0.0 0 5 10 1/normalised flow Figure 7. Relationship between inverse of normalised flow, and OP concentration relative to mean OP, for June to September. Averaged data by flow class for sites with > 100 data points. The flow on the groundwater dominated sites was less variable, and extremely low flows (less than about 20% of mean) were rare. SID 5 (Rev. 3/06) Page 19 of 29 Conclusions The objective was to quantify the impact of reduced water availability (due to agriculture) on water quality under present land use and climate. The underlying approach was to correlate quality measurements with river flow, to inform the assessment of the likely effects of reduced flow due to water abstraction for agricultural use. Reduction in flow due to abstraction of water will affect concentrations of pollutants either by reducing the quantity of water available for dilution of a downstream pollution source; or by changing in-river processes. Diffuse water pollution occurs mainly in the winter (when flows are at their highest). Point source pollution occurs all year, and at specific locations. We could therefore hypothesise that effects of water abstraction due to dilution effects will be greatest in the summer, and that effects will mainly be restricted to water quality parameters associated with point source inputs. By analysing GQA and HMS data, we were able to demonstrate: • The importance of point sources – the inverse relationship between flow and water quality parameters (i.e. reduced flow = increased concentration) was demonstrable for pollutants associated with sewage inputs (e.g. orthophosphate, chloride, boron). Concentrations of pollutants derived mainly from diffuse sources (e.g. nitrate, sediment) did not increase during summer, and were not greater in years of low summer flow. In general, these pollutants showed decreased concentrations during these periods. • Regional effects – the inverse relationship between flow and water quality parameters was stronger in regions with predominantly low rainfall and high populations (e.g. London, South East and East) and poorer in wetter areas of low population (e.g. South West, Wales). This again supports the assertion that effects are dominantly due to sewage related pollutants. • Groundwater-fed rivers – Flows tended to be at their lowest in July and August, apart from in Thames and Southern regions, where it was August and September. These latter regions have a greater proportion of groundwater-fed rivers where the groundwater sustains the surface water flows longer through the summer. We were also able to demonstrate different quantitative relationships between OP concentrations and flow for surface fed rivers compared with those that were predominantly groundwater-fed (see below). To summarise, flows are least in the summer months (which coincides with the period of irrigation demand). However, the effects of reduced flows (and, by inference, effects of any practices that abstract water at these times) vary regionally and very with type of pollutant. Effects appear to be restricted to sewage pollutants (and, again, by inference, any other pint pollutant source that feeds water systems during the summer months). Effects are greatest in summer and in the driest parts of the country – which are also the areas and times with the greatest irrigation demand. We have been able to develop quantitative relationships which provide an indication of the likely effects of reduced flow on water quality. A decrease in summer average river flow of 20% was estimated to result in an increase in the average chemistry classification of between 0.10 and 0.20 grade points. As the scale of the th chemistry classification thresholds is approximately linear, this corresponds to an increase in the 90 percentile -1 th BOD concentrations of 0.2 to 0.4 mg L and a decrease in the 10 percentile dissolved oxygen concentration of 1 to 2% across all rivers in a region. For orthophosphate, at average flow, concentration would be about 0.9 times the mean concentration; at half average flow, about 1.08 times the mean; and at 0.1 times normalised flow, concentrations will be about 2.5 times the mean. For surface water sites halving flow from 0.2 to 0.1 times mean flow, a change typical of summer low flows, would cause an increase in OP concentration from about 1.4 to 2.1 times mean OP concentration, an increase of about 50%. The flow on the groundwater dominated sites was less variable, and extremely low flows (less than about 20% of mean) were rare. On groundwater-dominated stream sites, halving of flow from 0.5 to 0.25 of average flow would increase OP concentrations from about 1.2 times the mean to 1.6 times the mean OP value, an increase of about one third. Equivalent relationships would be expected for other pollutants derived from point sources. While clear relationships were apparent for chloride and conductivity, data for other pollutants was more variable, and the regression equations explained less of the variance. However, as indicated both by the analysis for OP above and by theoretical analysis, abstraction of water during summer would increase the risk of excessively elevated concentrations of these pollutants. 3.5. Climate change effects (Objective 4) The objective was to provide a model based quantification of the surface water available for abstraction within the limits of good ecological status, and the proportion of this that is accounted for by current agricultural demand SID 5 (Rev. 3/06) Page 20 of 29 under present day and future climate scenarios. The general approach was to first develop a catchment scale water balance and river routing model to simulate daily river flows. Permitted levels of abstraction for good ecological status were then applied to the simulated river flows to calculate the maximum available water in the summer months and across the year, and were compared to mapped agricultural demand. Simulation of Natural river flow This project was dependent on the simulation of daily river flows for each of the 7,800 river catchments defined by the EA as part of the WFD catchment characterisation phase. In common with other studies, this required an approach to model application without calibration of parameters for individual catchments. Model parameters were derived from summary indices of present day catchment hydro-geological characteristics. The general approach was to use a daily soil moisture balance model to estimate effective rainfall, which was then partitioned into quick and slow response pathways and routed to give the catchment outflow on a daily basis. The parameters of this model were determined from present day calibration procedures and then held constant under changing rainfall and temperature time series for climate change scenarios. The catchment routing model was based on versions of the ‘Probability Distributed Model’ (PDM; Moore, 1985) and the ‘Identification of unit Hydrographs And Component flows from Rainfall, Evaporation and Streamflow’ model (IHACRES; Sefton & Boorman, 1997) in which daily time series of effective rainfall for a catchment unit are partitioned into two parallel reservoirs representing quick and slow flow. Effective rainfall was simulated using a modified form of the Daily Soil Moisture Accounting (DSMA) model described by Holmes et al. (2002). The DSMA model was calibrated with PETG data produced by the MORECS model (Hough et al., 1996). Complete details of the modelling approach are provided in Annexe I. Model validation Prior to application of the effective rainfall and river routing models to all catchments, they were applied to a number of validation catchments for which spatially averaged rainfall, temperature and observed river flows were available. Catchments classified as natural, i.e. the influence of abstractions or discharges is a maximum of 10%, were selected, which resulted in a dataset of 23 catchments. The fractional extents of soils of each LAND-HOST 2 class within each catchment were derived by intersecting soils and land cover data on a regular 1 by 1 km grid with the catchment boundaries. The land cover data were derived from the Land Cover Map of Great Britain (Fuller et al., 2002) and the MAGPIE database (Lord & Anthony, 2000). Model performance was assessed by evaluation of the bias in the annual average river flow and the NashSutcliffe criterion (Nash & Sutcliffe, 1970) which is sensitive to errors in the timing and magnitude of flow peaks. Simulated and measured annual average river flows were well correlated but were frequently under-estimated (average bias of 15%) although the original DSMA model had an average bias of only 2% (Fig. 8). Other model formulations such as IHACRES introduce a non-linear rainfall runoff model to correct this bias on a catchment by catchment basis. We did not correct the bias in this study as it was not clear whether the bias occurred in the estimation of the PETG or critical RC values for the selected catchments, or whether the bias was simply a result of a relatively short validation period. 2000 Simulated flow (mm) 1500 1000 500 0 0 500 1000 1500 2000 Observed flow (mm) Figure 8. Comparison of modelled and observed annual average river flows for selected river catchments using data for the period 1980-90. Values of the Nash-Sutcliffe criterion less than zero indicate a model that is less efficient than simply using the average observed flow as a predictor of the daily flow, whilst a value of one indicates a perfect model. The NashSutcliffe criterion for the daily flow simulations were generally in the range 0.30 to 0.91, with three outlier and very poor simulations of -0.7,0.06 and 0.17 for the Dun (39028), Tillingbourne (39029) and Foston Beck (26003) SID 5 (Rev. 3/06) Page 21 of 29 catchments. The criterion for the monthly simulations relevant to this study are generally higher and, excluding the three outliers, in the range of 0.53 to 0.95 indicating an overall good model fit to the validation data that is comparable to criterion in the range 0.45 to 0.83 reported by Arnell & Reynard (1996) also for a climate change impact study. The model performed well for flashy and groundwater dominated catchments, and for catchments with various levels of flow (more detail in Annexe I). The model was able to reproduce the year to year variation in observed river flows for catchments with both high (e.g. 60006) and low annual rainfall (e.g. 25005). This gave us some confidence that the model simulated the variation in annual AET correctly, in response to changes in available water, so that the model could be expected to be capable of simulating the impacts of long-term changes in rainfall as predicted for future climate. Climate scenarios Daily baseline and climate change scenarios on a 5 km grid across England and Wales for 2020s (2010-2039) and 2050 (2040-2069) low and high emission scenarios were generated from the Met. Office 1961-90 monthly data and the UKCIP02 monthly data by ADAS using the LARS-WG stochastic weather generator (Semenov & Barrow, 1997). The standard procedure when employing climate change scenarios is to use the low and high emission scenarios, as this was considered adequate to frame the extent of the problem. Definition of ‘available water’ for abstraction Standards for permitted levels of abstraction from surface water complying with good ecological status were defined by UKTAG (2006) and were summarised earlier. Following these rules, permitted abstraction is between 7.5% and 35% of the daily flow and depends on the river type (for a map of the river types see Fig. 1), season, and flow level. These rules were here applied to the simulated flow time series to estimate the surface water which is available for abstraction following these standards. Note that the definition is total available water, and takes no account of competing demands on this water, for example for public water supply. Results The simulated values of annual natural flow in England and Wales vary between 60 mm in the driest parts of East Anglia to nearly 3000 mm in Wales and the Northeast of England (Fig. 9), with the spatial variation closely following the variation of rainfall. Under future climate projections, annual flows decrease on average over England and Wales by between 2% to 5% for the 2020s low to the 2050s high emission scenario (Fig. 10). These changes in annual amounts appear relatively small and are because of the interactions between potential and actual evaporo-transpiration (PET and AET) and drier summers/wetter winters. The higher temperatures in the 2020s and 2050s result in higher PET but, due to reduced water availability especially in summer, AET remains nearly constant or decreases slightly. 1000 Base LO2020 900 HI2020 800 LO2050 HI2050 700 mm 600 500 400 300 200 100 0 rain Figure 9. Average annual simulated runoff for 1961-1990 baseline scenario for England and Wales. pet aet her Figure 10. Simulated annual water balance for 1961-1990 baseline and climate scenarios for the 2020s and 2050s, averaged over England and Wales. The national average change masks greater regional spatial variation. For example, in the 2050s high emission scenario a small increase in runoff is predicted of up to 15% in a narrow band on the East and South coast. More generally there is a 5 to 10% reduction in runoff. The spatial variation in the change in runoff (Fig. 11) is broadly similar to the change in rainfall and further differentiated by the spatial distribution of changes in AET. In the wettest parts of the country – e.g. mid Wales and the Northeast - AET increases, but it decreases in the drier parts of the country. This is caused by the decrease in annual precipitation and a more pronounced seasonality; AET is more limited in the drier summers, but does not increase as much in wetter winters as PET levels in winter SID 5 (Rev. 3/06) Page 22 of 29 are generally low. The reduction in summer rainfall also results in a larger decrease in river flows during the irrigation season from June to August with an average decrease for the 2050s high scenario over England and Wales by around 27% (Fig. 12). The greatest percentage reductions in summer flow occur in catchments with flashy river regimes. The smallest reductions occur in catchments dominated by baseflow, in which summer flows are partially maintained by the increased winter rainfall, which enters the groundwater store and is released in the summer months. Figure 11. Percentage change in yearly runoff for the 2050s high emission scenario. Figure 12. Percentage change in summer flow (Jun – Aug) for the 2050s high emission scenario. The range of changes in monthly flow is comparable to other studies (Arnell, 2004; Young & Reynard 2004; Walsh & Kilsby 2007). However, there can be great differences between different hydrological models, (e.g. 20% and more), in addition to uncertainties related to the climate projections (Prudhomme et al., 2005). The average yearly available water in the 1961-1990 baseline scenario was just under 100 mm on average over England and Wales but there was a large spatial variation and in c. 20% of East Anglia the yearly available water is less than 30 mm. In relation to abstraction for irrigation, it is not only the average yearly available water but, in particular, water available during the irrigation season between June and August and the variation between wet and dry years which are of concern. Averaged over England and Wales, the summer available water, defined as the water available between June to August inclusive, was around 7 mm (c. 7% of the yearly available water, Fig. 13) and varied between 1 to 100 mm. 100 8 average year dry year 80 Summer available water (mm) Yearly available water (mm) 7 60 40 20 6 5 4 3 2 1 0 0 Base LO2020 HI2020 LO2050 HI2050 Base LO2020 HI2020 LO2050 HI2050 Figure 13. Yearly and summer (June - August) available water in an average year, and the yearly and summer available water exceeded in 80% of the years, averaged over England and Wales for the 1961-1990 baseline scenario and for the 2020s and 2050s low and high emission scenarios. For the future scenarios, the average annual available water is only predicted to decrease by 1% for the 2020s low emission scenario to 4% for the 2050s high emission scenario, suggesting that for the annual available water the variation between years is more relevant than the long-term changes in future climate. Summer available water in an average year averaged over England and Wales decreases by 8% to 28% for the 2020s low to the 2050s high emission scenario. The changes, however, show considerable spatial variation (Figs 14 and 15) in England and Wales and the greatest percentage changes occur in the Western part of the country, where agricultural water demands are currently relatively low. The reduction in summer available water in a dry year is 7 to 18%. SID 5 (Rev. 3/06) Page 23 of 29 Figure 14. Percentage change of the yearly available water for 2050 high emission scenario compared to the 1961-1990 baseline scenario in England and Wales. Figure 15. Percentage change of the summer available water for 2050 high emission scenario compared to the 1961-1990 baseline scenario in England and Wales. Compared to the average yearly available water, the agricultural demand is relatively small so that virtually nowhere in England and Wales does the agricultural demand exceed the average yearly available water. However, for the summer months from June to August, there are a number of catchments, where the agricultural demand is greater than the summer available water. These catchments are mainly located in East Anglia, where the summer demand exceeds the summer available water in c. 30% of the area (Fig. 16). This area is extended in a year with a dry summer (defined as a year with a summer flow which is exceeded in 80% of the years), when the agricultural demand exceeds the available water in c. 60% of East Anglia (Fig. 17). In the areas where the demand is in excess of the available water, the demand averages 260% of the available water in an average year and 440% in a dry year. Figure 16. Average summer (Jun - Aug) available water minus 2001 agricultural summer demand. Figure 17. Summer (Jun - Aug) available water exceeded in 80% of the years minus summer agricultural demand for 2001 if a dry year. Assuming present day demands, the area where summer agricultural demands in an average year in the 2050s high emission scenario exceed summer available water from surface waters increases to 7% of England and Wales and 32% of East Anglia, this is an increase of around 20% compared to the baseline. In a dry summer, agricultural demands exceed summer available water in 19% of England and Wales and 68% of East Anglia, an increase of around 10% compared to the baseline (Fig. 18 and Table 9). SID 5 (Rev. 3/06) Page 24 of 29 Table 9. Percentage of area where the irrigation demand exceeds the available surface water during the summer months for an average year and a ‘dry year’ for the 1961-90 baseline scenario and the 2050s high emission scenario for England and Wales and for the EA Anglian Region. Baseline scenario Average summer Dry summer England & Wales East Anglia 6 26 HI2050s scenario Average summer Dry summer 18 62 7 32 England and Wales East Anglia 10000 Demand as percentage of available water 10000 Demand as percentage of available water 19 68 baseline average summer baseline dry summer 1000 2050s high average summer 2050s high dry summer 100 10 1 1000 100 baseline average summer 10 baseline dry summer 2050s high average summer 2050s high dry summer 1 0.1 0.1 0 20 40 60 80 100 0 20 40 60 80 100 Percentage area Percentage area Figure 18. Summer demand as percentage of available water for an average year and a ‘dry year’ for the 196190 baseline scenario and the 2050s high emission scenario for England and Wales and for the EA Anglian Region. Conclusions The developed model was adequate at simulating flow in a range of catchments, which allowed us to take this forward to simulate baseline flows (see earlier) and also the effects of selected climate change scenarios. Under future climate projections, it was calculated that annual flows would decrease on average over England and Wales by between 2% to 5% for the 2020s low to the 2050s high emission scenarios, respectively. These changes in annual amounts appear relatively small and are because of the interactions between potential and actual evaporo-transpiration (PET and AET) and drier summers/wetter winters. The higher temperatures in the 2020s and 2050s result in higher PET but, due to reduced water availability especially in summer, AET remains nearly constant or decreases slightly. The reduction in summer rainfall however results in a larger decrease in river flows during the irrigation season from June to August with an average decrease for the 2050s high scenario over England and Wales by around 27%. The simulations also show that the greatest percentage reductions in summer flow occur in catchments with flashy river regimes. The smallest reductions occur in catchments dominated by baseflow, in which summer flows are partially maintained by the increased winter rainfall, which enters the groundwater store and is released in the summer months. We know from water use statistics that agriculture uses only about 1% of abstracted water and it is therefore not surprising that, when considering annual water availability and agricultural demand, there are no conflicts demonstrated by the model simulations. As described earlier, it is the summer months that are critical for 6 3 irrigation requirement and this is when most of the 125 x 10 m is applied (plus livestock/nursery requirements). Repeating this analysis of comparing available water and agricultural demand for the summer months only, shows that (for baseline climate conditions), about 30% of East Anglia (and about 7% of England and Wales) is affected. The area where demand is greater than availability increases to about 60% (East Anglia) or 20% (England and Wales) in a dry year (defined as a year with 80% of average flow). The difference in area affected between a wet and a dry year is much greater than the difference between baseline and the 2050 high scenario. We can conclude, therefore, that the short-term annual variations in weather have a greater influence than longterm trends. SID 5 (Rev. 3/06) Page 25 of 29 In terms of water resource planning, however, the climatic changes are important, in that irrigation systems and abstraction licensing must be designed to function in the ‘design dry year’ rather than an average year. The economic benefit of the system is realised largely in the driest years. It is important to note that the simulations take no account of increased irrigation demand due to climate change. However, it is possible that the irrigation demands will increase. Irrigation demand in a dry year is predicted to increase by around 20% in the 2020s and 30% in the 2050s (Downing et al., 2003). These increases vary spatially and are greatest in the South and East and less in the West and North. The change in demand therefore further worsens the situation in East Anglia. An increase in the demand of 30% would mean that in a dry summer the area where summer agricultural demands exceed the available water increases to 24% of England and Wales and 73% of East Anglia. Finally, irrigation demand overall may increase under climate change, due to a combination of: • Extension of irrigation to a wider range of crops, due to the increased frequency or severity of drought. For example, maize is a major irrigated crop in many countries. • Introduction of new crops which may require irrigation. 3.6. Regional impacts (Objective 5) The aim was to review the regional pattern of impact across England and Wales, with special attention to differences between types of pollutants. The spatial analysis and modelling has demonstrated that: • On an annual basis, there is little conflict between water available in a catchment and that required for agriculture. However, there is more pressure in the East than the West. • Water quality/chemistry – the impacts of reduced flow are most clearly demonstrable where there is a significant input of sewage treatment (i.e. areas of high population) and low flows (i.e. drier areas). Therefore, effects are greatest in the east and south east. Low flows could, of course, directly affect other parameters, such as oxygen status, in all areas. • Although water demand for livestock is greater than for irrigation (in total): o o • Livestock are predominantly in the wetter west Irrigation is restricted to the June-August period and is mainly required in the east If we focus on this summer period, when flows will be at their lowest anyway: o o The greatest pressure on water resources is in the east, with large proportions of East Anglia showing a shortfall between irrigation demand and calculated available water. There is an increase in the concentration of sewage-derived pollutants, but not nitrate and sediment (more associated with diffuse pollution. • There is evidence that groundwater-fed rivers (predominantly in the south and south-east) reach their lowest flows in August/September, somewhat later than surface water-fed rivers (July/August). This means that at least part of the main irrigation demand commonly occurs before the lowest flows. Climate change also has less impact on summer flows in groundwater-fed rivers because the increased winter rainfall is able to replenish the reserves that sustain flow. • Overall, our predictions show that the total effect of climate change would be in the range of a 2-5% reduction in run-off. Effects would be greatest in the wetter regions because summer AET will be closer to PET whereas in the east PET is limited by the reduced summer rainfall under climate change. Nevertheless, climate change will increase the pressure on water resources in the east. It is interesting to note that we have demonstrated greater variation in run-off between years than we are predicting in change between the average baseline and future climate change. 4. CONCLUSIONS 4.1. Consequences for the Water Framework Directive (Objective 6) This project has tried to assess the quantity of water available for abstraction from surface waters for the current (baseline) climate, as well as estimating effects under future climate scenarios. We have been able to place these in the context of the agricultural demand for water. SID 5 (Rev. 3/06) Page 26 of 29 A model was developed for the simulation of time series of flow for the whole of England and Wales. The model is responsive to catchment characteristics of land use and hydrogeology and it showed a good performance over a range of different catchments, without the need for site-specific calibration to observed data. The evaluation of agricultural demands in the context of available water showed that demands from agriculture are significant in the summer period and in certain areas of the country, mainly concentrated in East Anglia. Also, agriculture is presumably the most important abstractor in headwaters, as other abstractors usually abstract from major rivers. In many of the areas in which abstraction of summer surface water is classified as unsustainable by the EA, agricultural demands are not significant. Abstraction of summer surface water is currently considered unsustainable in all of the Southern Region and parts of East Anglia (Fig. 19), but in most parts of the Southern Region agricultural demands are less than 3 mm. This is because the crops requiring high amounts of irrigation are not grown to a big extent in this region. In East Anglia, where irrigation demands are significant, summer surface water abstractions are classified as unsustainable in smaller parts, and no additional water is available in any part of East Anglia. With decreasing summer flows under future climate conditions, less water will be available for abstractions, while demands are expected to increase. Figure 19. Areas of unsustainable surface water abstraction as defined by the Environment Agency (2001) for summer surface water abstractions (1) and winter surface water abstractions (2). In terms of water quality, reassessment of standards under the WFD has created closer linkages between chemical, nutrient and hydromporphological status and ecology. Thus, if agricultural water demand impacts on dissolved oxygen, BOD, ammonia or P levels and/or flow, then this could have potential impacts on the ecological status. We have concluded that changes to water quality due to dilution effects are: • chiefly related to point sources (sewage treatment works) rather than diffuse agricultural sources; • are is most likely in the east; • are most likely in the summer months. In terms of climate change impacts on water availability, and the demand for irrigation water in summer, we concluded that: • Climate change is likely to increase demand for irrigation water during summer, and increase the frequency of occurrence of extreme demand years and of low flows; • annual variations are as great as or greater than expected changes in mean value within the next few decades However the change in mean value, and in frequency of demand exceeding availability, will impact on water resource planning and irrigation abstraction licensing. This analysis did not take explicit account of other potential effects of climate change. Greater risk of summer drought, with higher temperatures, could increase irrigation demand: • by extension of Irrigation to a wider range of crops; • by introduction of new crops requiring irrigation; SID 5 (Rev. 3/06) Page 27 of 29 • by extension of irrigation into areas where irrigated crops are currently rare – i.e. towards the west. The extension of irrigated crops towards areas where they currently do not occur could have adverse impact on rivers which are currently of relatively good ecological quality: • by reduction of summer flows; • by quality changes as discussed above; • more directly, by the fact that some of these crops carry with them an increased risk of sediment or other pollutant transfer. 4.2 Recommendations The results form a basis of information for input to river basin planning, and for taking account of impacts of climate change, in a uniform way. They should be made available to river basin planning teams. The results suggest that conflict between supply and demand is greatest during summer, and in the east; and that this conflict will become greater rather than less. This suggests that, in future, irrigation may only be viable with winter storage to allow abstraction of water while flows are high. Further assessment of the impacts of climate change should take account of likely changes in land use, and in crops for which irrigation may become economic. As a result of climate change and consequent land management change, and the difficulties of irrigation in eastern areas, it is possible that irrigation will increase in more sensitive western catchments. This may be a result of changed cropping (greater areas of responsive crops grown in the west) and/or a need to irrigate crops in the west due to the changed climate. The risk of this, and the ecological implications, should be investigated. References to published material 9. This section should be used to record links (hypertext links where possible) or references to other published material generated by, or relating to this project. SID 5 (Rev. 3/06) Page 28 of 29 Acreman, M. & Dunbar, M.J. (2004). Defining environmental river flow requirements – a review. Hydrology and Earth Sciences 8, 861-876. Arnell (2004) Relative effects of multi-decadal climatic variability and changes in the mean and variability of climate due to global warming: future streamflows in Britain. Journal of Hydrology 270, 195–213. Arnell, N.W. & Reynard, N.S. (1996). The effects of climate change due to global warming on river flows in Great Britain. Journal of Hydrology 183, 397–424. Bailey, R.J. (1990). Irrigated crops and their management. Farming Press, Ipswich. Carvalho, L., Bennion H., Darwell A., Gunn, I., Lyle, A., Monteith, D., & Wade, M. (2003). Physico-chemical conditions for supporting different levels of biological quantity for the Water Framework Directive for freshwaters. Report by the Environment Agency. Defra (2005) e-Digest of Environmental Statistics, Published December 2005. http://www.defra.gov.uk/environment/statistics/index.htm Defra (2007). The Protection of Waters Against Pollution from Agriculture, Consultation on implementation of the Nitartes Directive in England. August 2007. www.defra.gov.uk/...waterpollution-nitrates/consultation.pdf Downing, T.E., Butterfield, R.E., Edmonds, B., Knox, J.W., Moss, S., Piper, B.S. & Weatherhead, E.K. (2003). Climate Change and the Demand for Water, Research Report, Stockholm Environment Institute Oxford Office, Oxford. 201pp. Deflandre, A. & Jarvie, H. (2004). Nutrients causing Eutrophication. In REBECCA WP4 Rivers Relationships between ecological and chemical status in surface waters. EU Sixth Framework Programme project. Contract No. SSPI-CT-2003-502158 Defra (2000). Economic instruments in relation to water abstraction. http://www.defra.gov.uk/environment/water/resources/econinst/ Dyson, M., Bergkamp, G. & Scanlon, J. (Eds) (2003). Flow: essentials of environmental flows. IUCN, Gland, Switzerland and Cambridge, UK. Environment Agency. (2002). Managing Water Abstraction. The Catchment Abstraction Management Strategy process. Environment Agency. Fuller, R.M., Smith, G.M., Sanderson, J.M., Hill, R.A. & Thomson, A.G. (2002). The UK Land Cover Map 2000: Construction of a parcel-based vector Map from satellite images. Cartographic Journal 39, 15-25. Jarvie, H.P., Neal, C. & Withers, P.J.A. (2006). Sewage-effluent phosphorus: A greater risk to river Eutrophication than agricultural phosphorus? The Science of the Total Environment 360, 246– 253. Holmes, M.G.R., Young, A.R., Gustard, A. & Grew, R. (2002). A new approach to estimating mean flow in the UK. Hydrology and Earth System Sciences. 6(4), 709-720.. Hough, M., Palmer, S., Weir, A., Lee, M. & Barrie, I. (1996). The Meteorological Office Rainfall and Evaporation Calculation System: MORECS Version 2.0 (1995). An update to Hydrological Memorandum 45. Meteorological Office, Bracknell, pp.80. Hulme, M., Briffa, K.R., Jones, P.D. & Senior, C.A. (1993). Validation of GCM Control Simulations Using Indices of Daily Airflow Types over the British Isles. Climate Dynamics 9, 95–105. King, J., Tiffin, D., Drakes, D. & Smith, K.A. (2006). Water use in agriculture: establishing a baseline. Report for Defra Project WU0102. www.defra.gov.uk/science/project_data/DocumentLibrary/. Lord, E.I. & Anthony, S.G. (2000). MAGPIE: A modelling framework for evaluating nitrate losses at national and catchment scales. Soil Use and Management 16, 167-174. Moore, R.J. (1985). The probability-distributed principle and runoff production at point and basin scales. Hydrological Science Journal 30, 273-297. Nash, J.E. & Sutcliffe, J.V. (1970). River flow forecasting through conceptual models, Part 1 - A discussion of principles. Journal of Hydrology 10, 238-250. OFWAT (2006). Security of supply, leakage and water efficiency: 2005-06 report . www.ofwat.gov.uk/aptrix/ofwat/publish.nsf Prudhomme, C., Piper, B., Osborn, T. & Davies, H. (2005). Climate change uncertainty in water resource planning. Report to UKWIR, UK. 305 pp. Sefton, C.E.M & Boorman, D.B. (1997). A regional investigation of climate change impacts on UK streamflows. Journal of Hydrology 195, 26-44. Semenov, M.A., & Barrow, E.M. (1997). Use of a stochastic weather generator in the development of climate change scenarios. Climate Change 35, 397–414. Thompson, A.J., King, J., Smith, K.A. & Tiffin, D.H. (2007). Opportunities for reducing water use in Agriculture. Defra Project WU0101. www.defra.gov.uk/science/project_data/DocumentLibrary/. UKTAG. (2006). UK Environmental Standards and Conditions ( PHASE 1). Final Report (SR1-2006). UK Technical Advisory Group on the Water Framework Directive. Walsh, C.L. & Kilsb,y C.G. (2007). Implications of climate change on flow regime affecting Atlantic salmon. Hydrology and Earth System Science. 11(3), 1127-1143. Weatherhead E.K., Knox, J.W., Morris, J., Hess, T.M., Bradley, R.I. & Sanders, C.L. (1997). Irrigation demand and on-farm water conservation in England and Wales. Report to Ministry of Agriculture, Fisheries and Food (MAFF), Project OC9219 Weatherhead, E.K. & Danert, K. (2002). Survey of irrigation of outdoor crops in 2001 - England. Institute of Water and Environment, Cranfield University. Weatherhead, E.K. (2007). Survey of irrigation of outdoor crops in 2005 - England and Wales. Institute of Water and Environment, Cranfield University. Weatherhead, E.K., Place, A.J., Morris, J. & Burton, M. (1994). Demand for Irrigation Water. NRA R&D Report 14. HMSO: London. Young, A.R. & Reynard, N.S. (2004). River Flow Simulation within Ungauged Catchments; the Utility of Regionalised Models. In: Complexity and Integrated Resources Management, Transactions of the 2nd Biennial Meeting of the iEMSs (Edited by Claudia Pahl-Wostl, Sonja Schmidt, Andrea E. Rizzoli, and Anthony J. Jakeman). 6 pp. SID 5 (Rev. 3/06) Page 29 of 29