Article ID: | iaor20124021 |
Volume: | 56 |
Issue: | 1-2 |
Start Page Number: | 1 |
End Page Number: | 13 |
Publication Date: | Jul 2012 |
Journal: | Mathematical and Computer Modelling |
Authors: | Peng Shizhang, Luo Yufeng, Khan Shahbaz, Rana Tariq, Wang Weiguang, Jiang Yunlu |
Keywords: | simulation, statistics: inference |
The estimation of groundwater evapotranspiration (ET) helps evaluate the risk of soil salinisation and the capacity of regional water resources. This paper focused on the effect of discretisation cell size on the output uncertainty of regional groundwater evapotranspiration modelling. The study area, the Liuyuankou Irrigation System (LIS), was discretised into cell sizes of 100×100 m, 200×200 m, 500×500 m, 1000×1000 m, and 2000×2000 m. Digital elevation models (DEM) at the lower resolutions were obtained by resampling the 90 m Shuttle Radar Topography Mission (SRTM) DEM using an averaging algorithm. The krigged groundwater table was subtracted from the ground surface elevation to generate the groundwater depth maps for various cell sizes. For each resolution, the groundwater ET was calculated using two ET functions (linear and exponential) for each cell. Lastly, the effects of the cell size on the output of the regional groundwater ET modelling were evaluated. The results showed that lower ET rates were obtained when a coarser cell size was used due to the smothering of the surface elevations and the groundwater depths, regardless of the employed ET functions. From comparisons of the delineated area, the cell numbers, the groundwater depth maps and the simulated groundwater ET rates, a discretisation cell size of 500 m was recommended for the LIS to balance the model’s accuracy and computation efficiency in groundwater ET modelling.