The hydrological impacts of energy crop production in the UK. Final report

Finch, J. W.; Hall, R. L.; Rosier, P. T. W.; Clark, D. B. ORCID:; Stratford, C. ORCID:; Davies, H. N.; Marsh, T. J.; Roberts, J. M.; Riche, A.; Christian, D.. 2004 The hydrological impacts of energy crop production in the UK. Final report. London, UK, Department of Trade and Industry, 151pp. (CEH Project Number: C01937, B/CR/000783/00/00) (Unpublished)

Before downloading, please read NORA policies.

Download (26MB)


This report describes the work carried out between March 2002 and January 2004 under ETSU Contract number B/CR/000783/00/00 by the Centre for Ecology and Hydrology, Wallingford. It also describes the results of measurements made by Rothamsted Research staff under a sub-contract. The objectives of this work are: 1. To determine the effects on water availability at the catchment and sub-catchment scale, of production of energy crops, across England and Wales. 2. To indicate areas where the crops will be most productive, which will be largely determined by water availability. The two objectives have been met by a programme of field measurements, to provide parameter values for a numerical model and data to test the predictions of the model, and a numerical model that produced predictions of the water use of the current land cover and four energy crops: willow short rotation coppice, poplar short rotation coppice, Miscanthus and switchgrass, for a wet, a dry and a typical year. A demonstration GIS has been produced which shows how the predictions can be made use of in catchment management. Measurements The purpose of the measurements was to provide values for parameters used by numerical models and data against which the predictions of the model could be tested. Values for some parameters were available from the literature or previous studies A previous study had made measurements on poplar short rotation coppices (SRC) and so field measurements took place on willow SRC at Roves Farm (Wiltshire) and Miscanthus and switchgrass in experimental plots at Rothamsted and Woburn (Bedfordshire) and a field of Miscanthus near Richard’s Castle (Herefordshire). Willow SRC Measurements of the stomatal conductance were made in order to develop a relationship between the stomatal conductance and the atmospheric drivers on evaporative demand. These measurements were compatible with a published relationship. The evaporative flux of the willow SRC was measured using the eddy correlation technique, applied to data obtained from a sonic anemometer, and provided data for model testing. This was also the purpose of sap flow gauges which were used to quantify the transpiration of individual stools. These data were scaled up to the full canopy on the basis of measurements of the number of stools and the leaf area. The enhanced evaporation at the edges of plantations was investigated using the transpiration data and measurements of the soil water content in the top 10 cm. The results showed that the effect was only significant at the extreme edge of the coppice and is the result of the larger leaf area occurring at the edge. A simple relationship has been developed to capture this effect. Miscanthus and switchgrass Measurements of the stomatal conductance enabled a model of the variation of the stomatal conductance of the switchgrass as a result of the atmospheric drivers to be calibrated. Measurements of the interception by switchgrass were made during two seasons. These data were used to quantify the canopy capacity and relate it to the leaf area index. Measurements of the changes in the soil water with depth under both Miscanthus and switchgrass were made to enable the numerical model to be tested. For the same purpose, measurements were made of the evaporative flux from Miscanthus at the site near Richard’s Castle. Numerical modelling The numerical model used was based on the Met. Office Surface Energy Scheme (MOSES) which is physically based and provides a comprehensive description of the exchanges of energy, water and carbon at the land surface. The data collected on this project were used to give values for the parameters that determine the model’s description of the energy and water balances. The results of running the model to predict the water use of the energy crops were tested against measurements of soil water content and evaporation made in this project. The values of the parameters for the carbon balance of the MOSES were determined by calibrating the model against the measured canopy height and leaf area index. An analysis of the rainfall data for England and Wales enabled the choice of dry, wet and typical years to be made, viz: 1976, 1988 and 1982 respectively. The analysis also demonstrated that the water year (the period running from the beginning of October in the previous year to the end of September in the year) was appropriate, given the importance of the amount of the soil water stored in the period prior to the growing season. The MOSES model was run on grid cells of 1 km2. The fractions of existing land cover classes (broadleaf woodland, needliferous woodland, grassland, shrubs, tilled land, urban, bare soil and water) were determined by aggregating the 25 classes of the Land Cover Map 2000 to eight MOSES classes and then calculating the fractions in each model grid cell. The soil hydraulic properties were obtained using the Hydrology Of Soil Types (HOST) as the basis. The dominant HOST class of each 1 km2 was used. The properties were obtained by producing the average texture of the soil series within a HOST class and then using pedo-transfer models to calculate the properties. The daily meteorological driving variables were obtained from the Met Office as values on a 1600 km2 grid. These were interpolated to the model grid using a bilinear algorithm. A further refinement was made to the values of air temperature by allowing for the altitude of the grid cell using the standard lapse rate. The greater spatial variability of the rainfall was handled by using a data set of monthly totals on a 1 km2 grid, which were disaggregated to daily values using the data from the 1600 km2 grid. There is no simple answer to whether energy crops will use more water than the existing land cover. It is a function of a number factors that include: the current type of the land cover; the specific energy crop, the amount of rainfall and the hydraulic properties of the soils. This was investigated in two ways: monthly time series for eight grid cells, to investigate the detailed temporal variability, and spatial distributions for water years with typical high and low rainfall, to investigate the spatial variability. Monthly time series The model was run on eight selected grid cells, representative of a range of climatological, soil and land cover conditions, to generate monthly times series, covering the period 1971 to 2000, of the predicted water use of each of the four energy crops, the change in water use from the existing land cover and the indicative yield. These time series allow the seasonal and inter-annual variability to be assessed. The results show that poplar SRC has a much higher transpiration rate than the other energy crops. This is due to the stomata having little response to high atmospheric evaporative demand. The effect of soils with low soil water storage in reducing indicative yield is reflected in the results. The three year harvesting cycle, assumed for the SRC, is apparent in the results and shows an increase in water use in successive years. The switchgrass and Miscanthus show an annual pattern that reflects the seasons but, there is a shorter growing season than for other land cover types because the threshold temperature, below which photosynthesis ceases, is higher Spatial distribution The predicted spatial distributions of the water use, change in water use and the indicative yield are presented as 39 maps which show: the predicted annual water use for the current land cover and the four energy crops, the change in annual water use if the current land cover were replaced with one of the energy crops and the indicative yield, all for water years with typical, high and low rainfall. The results show that, when soil water is plentiful, poplar SRC uses considerably more water that the current land cover or any of the other energy crops and also has significantly higher indicative yields. In areas of low rainfall, the rapid rate of transpiration early in the year rapidly depletes the soil water store with a result that growth is restricted in the later part of the year so that a low indicative yield is obtained. The conclusion is that the varieties of poplar SRC that the model was calibrated for are probably only viable in areas of high rainfall. Willow SRC has a higher water use than the existing land cover in most situations. This is mainly because it is specified in the model as having a greater rooting depth than the other land cover types, with the exception of broadleaf woodland and so is less affected by soil water stress. Miscanthus and switchgrass are predicted to have a lower water use than the existing land cover in most areas. This is because the canopy develops later in the year, due to these grasses being specified to have a higher threshold temperature below which photosynthesis ceases, with the result that the period when the transpiration occurs is shorter and occurs during the summer months. In addition, they have a higher water efficiency than the other land cover types because they use the C4 photosynthetic pathway. To some extent, these factors are offset by the rooting depth, which is intermediate between the existing woodlands and grasses. However, it is acknowledged that measurements through at least one full year would be needed to confirm this prediction. Demonstration GIS Software for a simple demonstration GIS has been written which, for a 60  60 km area of the head waters of the river Severn, allows the user to explore the possible impact of energy crop plantations on the water losses within a catchment. Conclusions • The effect of enhanced evaporation at the edges of SRC plantations is localised and so will have the greatest impact on small plantations. For plantation greater than 10 ha the effect is certainly comparable or less than other factors, e.g. the nature of the soil. • For the same rainfall and soils, the water use of the energy grasses is likely to be less or comparable to that of the existing land cover where it is grass or tilled land and less if the existing land cover is woodland or heathland. • In the final year of the three year cropping cycle, the water use of SRC is likely to be greater than the existing land cover if it is grass or tilled land and comparable or greater if it is woodland or heathland. However, in the first year it is likely to be less than existing land cover types. • The results for poplar SRC show a very high water use. These results should be interpreted with caution as it is likely that varieties could be or are available that would have lower water use. In which case the water use is likely to be comparable to that of willow SRC. • In areas of high annual average rainfall (greater than around 800 mm), the nature of the soil has little impact on the water use of the energy crops, or the existing land cover. However, in other areas, the soil hydraulic properties, particularly the ability to store water that can then be used by the plants for transpiration, can be important because of the higher transpiration rates of the energy crops. • When the rooting depth of the energy crop is deeper than the existing land cover, there is the possibility that, after a period of drought, the soil water deficits will be greater resulting in a reduction in recharge and/or runoff in the following winter. • During years with above average rainfall, when the transpiration rates are not constrained by soil water, the energy crops tend to use more water, than the existing land cover, mainly due to the higher interception losses. • The predicted indicative yield from the energy crops is a function of air temperature and the amount of sunshine. The energy grasses are predicted to be more sensitive to these factors than the SRC and so show a more marked trend of decreasing indicative yield with increasing latitude and altitude. • There are strong indications that, in areas of low annual average rainfall (less than about 700 mm), the indicative yields of all the energy crops are reduced by soil water stress.

Item Type: Publication - Report
Programmes: CEH Programmes pre-2009 publications > Water
UKCEH and CEH Sections/Science Areas: Harding (to July 2011)
_ Hydrological Risks & Resources
_ Process Hydrology
Funders/Sponsors: Department of Trade and Industry
NORA Subject Terms: Hydrology
Date made live: 09 Jan 2009 12:18 +0 (UTC)

Actions (login required)

View Item View Item

Document Downloads

Downloads for past 30 days

Downloads per month over past year

More statistics for this item...