19 November 2013

Plugging Water's Effects into an Earth System Model

Science

Connecting several data pipes in a popular land model, a research team led by Pacific Northwest National Laboratory simulated how irrigation from both surface water and groundwater affects the Earth’s water and energy budget. The surface water irrigation results varied depending on the observational sources used. Seasonal and year-to-year variations were large, and tended to be more pronounced in dry versus wet years. Even when the model was calibrated with agricultural observations, the difference between simulations showed a wide variability.

In a second study, the researchers introduced a method to represent irrigation from groundwater pumping for the first time. They found that groundwater-fed irrigation could lead to depletion of regional groundwater aquifers and unsustainable groundwater use in certain agricultural regions.

Approach

The team included a groundwater pumping design coupled to its irrigation module in the Community Land Model (CLM4) and configured it to simulate land surface water and energy budgets at a detailed resolution. They used two sets of data sources for irrigation area maps: 1) The Global Map of Irrigated Area (GMIA), a map that represents the fractional irrigated area around the year 2000, which combines reports from the Food and Agriculture Organization of the United National and the U.N. Ministries of Agriculture, and land-use and land-cover data from the U.S. Geological Survey; and 2) a high-resolution irrigated area map for the continental United States that combines the Nadir Bidirectional Distribution Function Adjusted Reflectance (NBAR) data during 2001, gridded climate-based indices of the surface moisture status, and a map of cultivated areas. With this data, they performed numerical experiments, with and without irrigation or groundwater pumping, to understand the impact of irrigation on terrestrial water cycling and the sensitivity of the model results to various input datasets, parameter values and irrigation water source options.

They found that at seasonal to inter-annual time scales, the effects of irrigation on the surface energy budget were large and persistent—more pronounced in dry years than wet years. Even with model calibration that produced overall good agreement with the irrigation amounts from the National Agricultural Statistics Service, differences between the two irrigation area datasets still dominated the differences in the inter-annual variability of land surface responses to irrigation. Their results also show that changes in soil moisture content induced by groundwater-fed irrigation have significantly altered the water availability and distribution, with large effects on land surface fluxes through surface-subsurface interactions.

Groundwater pumping can lead to fast depletion and unsustainable groundwater use in agricultural regions that have a low recharge rate (downward movement of surface water to the aquifer) and a deep groundwater table. Therefore, large-scale pumping should be included in Earth system models to depict the effects of irrigation on the regional water cycle.

Both studies highlight the challenges of depicting realistic irrigation effects on climate and the need to include a more complete representation of irrigation, surface and subsurface hydrology, and water management in Earth system models.

Impact

When the tap is turned, will there be water? As agricultural practices modernize and increase capacity around the world, agricultural irrigation is burdening water resources. World agriculture uses over 80 percent of fresh water sources, competing with other uses such as household, industrial and energy production. Irrigation and agricultural practices also have a substantial impact on the regional and local climate and the land’s surface water system through evaporation and plant transpiration. In these studies, the researchers incorporated impacts of irrigation from both surface sourced and groundwater pumping in a land surface model to better understand how it affects the Earth and changes to the atmosphere. This work is an important step to understanding historical and future climate change.

Summary

Leng et al. 2013. World agriculture consumes about 87 percent of global fresh water withdrawal and significantly impacts the global water cycle. Understanding the impact of irrigation on land surface fluxes, surface and subsurface states, and their interactions with atmospheric processes is crucial to understand historical climate change and model future climate at local and regional scales. Previous sensitivity studies of irrigation impacts on land surface fluxes/states show limited analysis of uncertainties from the input data and model irrigation schemes. A team of scientists, led by U.S. Department of Energy researchers at Pacific Northwest National Laboratory, improved the performance of the Community Land Model version 4 (CLM4) in simulating irrigation water use and surface fluxes by calibrating the model against data from agriculture census. They found that by using the irrigation area fraction datasets from two widely used sources as inputs, CLM4 tended to produce unrealistically large temporal variations of irrigation demand for applications at the water resources region scale over the conterminous United States. At seasonal to interannual time scales, the effects of irrigation on surface energy partitioning appears to be significant and persistent, and more pronounced in dry than wet years. Even with model calibration to yield overall good agreement with the irrigation amounts from the National Agricultural Statistics Service, differences between the two irrigation area data sets still dominated the differences in the interannual variability of land surface responses to irrigation. Their results suggest that CLM4-simulated irrigation amount and surface fluxes could be improved by calibrating model parameter values and accurately representing the spatial distribution and intensity of irrigated areas. The research also recommends a critical path forward to a realistic assessment of irrigation impacts using an earth system modeling approach by further developing CLM by including groundwater pumping and irrigation efficiency modules, and coupling CLM with streamflow routing and water management modules to account for all sources of water supply, which will be reported in follow-up studies. Leng et al. 2013. Human alteration of the land surface hydrologic cycle is substantial. Local water management practices, including groundwater pumping and irrigation, could significantly alter the quantity and distribution of water in the terrestrial system, with potential impacts on weather and climate through land-atmosphere feedbacks. A team of scientists, led by U.S. Department of Energy researchers at Pacific Northwest National Laboratory, incorporated a groundwater withdrawal scheme in the Community Land Model (CLM) and used the model to simulate the effects of groundwater-fed irrigation on local and regional water and energy budgets over the conterminous United States. Their results show that changes in the amount of soil moisture contributed by groundwater-fed irrigation significantly alters the water availability and distribution, with large effects on land surface fluxes through surface-subsurface interactions. Of the four regions analyzed in detail, only the Lower Mississippi region can support sustainable groundwater use because of the high recharge rate (downward movement of surface water to the aquifer) and shallow groundwater table. In contrast, low recharge rate combined with high groundwater withdrawal rate in the southern Great Plains, California, and Pacific Northwest lead to fast depletion of groundwater. As the groundwater table deepens, irrigation may become less effective in increasing soil moisture because more water administrated to the root zone percolates and recharges the groundwater. The research shows that large-scale pumping, currently omitted in most irrigation modeling studies, might greatly alter the hydrologic and land energy fluxes in agricultural regions, and further influence the local and regional climate. However, comparisons between simulations and observations from the U.S. Geological Survey and satellite data from the Gravity Recovery And Climate Experiment (GRACE) suggest that further development of the groundwater parameterization in CLM is needed to effectively represent the two-way interactions between surface water and groundwater, so that groundwater depletion due to groundwater-fed irrigation could be more realistically simulated.

Contact
Jeremy G Fyke
Los Alamos National Laboratory (LANL)
Acknowledgments

These studies were supported by the U.S. Department of Energy Office of Science Integrated Earth System Modeling (iESM) project for the Earth System Modeling and Integrated Assessment Research program. The Platform for Regional Integrated Modeling and Analysis (PRIMA) initiative provided the model configuration and datasets used in the numerical experiments. PRIMA is funded by PNNL's Laboratory Directed Research and Development Program.