Dataset used for the application of Budyko in irrigation areas

In this study, we evaluated the potential application of the Budyko hypothesis in agricultural irrigation areas. 8-year observed (2010-2017) data about 371 large irrigation districts including annual irrigation water, specified location, total irrigated area and irrigation water use efficiency were collected and provided by China Irrigation and Drainage Development Centre. Daily precipitation and monthly meteorological data including wind speed, air temperature, and relative humidity from weather stations on or around the selected irrigation districts covering the same period were downloaded from China meteorological data network (http://data.cma.cn/). The values of NDVI (Normalized Difference Vegetation Index) were extracted from MOD13A1 products with spatial-temporal resolution of 500 m and 16 d, which were available to download from the NASA Data Centre at https://reverb.echo.nasa. gov. The Digital Elevation Model (DEM) data with a spatial resolution of 1 km were downloaded from http://srtm.csi.cgiar.org/. The distribution map of soil texture denoting the proportion of sand and clay was provided by Data Centre for Resources and Environmental Sciences, Chinese Academy of Sciences (RESDC) (http://www.resdc.cn). The actual evapotranspiration was estimated as the sum of net irrigation water and effective precipitation by water balance equation. Irrigation activities have changed the natural hydrological processes and influence the allocation of water availability. By incorporating both irrigation water and precipitation in water availability, Budyko framework performed well in irrigation areas. In arid and semi-arid areas, 10% increase in irrigation water brings 1.58% to 5.29% increase in evapotranspiration, and 10% increase in precipitation brings 0.97% to 4.25% increase. For humid and semi-humid areas, the variation of evapotranspiration is mainly caused by energy supply and a 10% increase in potential evaporation brings 6.79% to 8.61% increase in evapotranspiration. Using the 8-year data, the optimal values of Budyko parameter ω was obtained by least square method. The values of ω in humid and semi-humid areas were generally large than those in arid and semi-arid areas. An empirical equation was developed to describe the obvious relationship between ω and NDVI and soil property. The equation performed well in reproducing parameter ω given its simplicity and easy accessibility to input factors.本研究评估了布迪科假说（Budyko hypothesis）在农业灌区的潜在应用价值。本研究采用由中国灌溉排水发展中心收集并提供的2010-2017年共8年的观测数据，涵盖371个大型灌区，数据包含年灌溉用水量、具体区位、总灌溉面积及灌溉水利用效率。同期覆盖所选灌区及其周边气象站的日降水量与月尺度气象数据（含风速、气温、相对湿度），则从中国气象数据网（http://data.cma.cn/）下载获取。归一化差分植被指数（Normalized Difference Vegetation Index，NDVI）数据从时空分辨率为500米、16天的MOD13A1产品中提取，该产品可从美国国家航空航天局数据中心（NASA Data Centre，https://reverb.echo.nasa.gov）下载获取。空间分辨率为1千米的数字高程模型（Digital Elevation Model，DEM）数据，从http://srtm.csi.cgiar.org/下载获得。表征砂粒与黏粒占比的土壤质地分布图，由中国科学院资源环境科学数据中心（Data Centre for Resources and Environmental Sciences, Chinese Academy of Sciences，RESDC，http://www.resdc.cn）提供。实际蒸散发量通过水量平衡方程，以净灌溉水量与有效降水量之和进行估算。 灌溉活动改变了自然水文过程，并影响水资源可利用量的分配。将灌溉用水与降水量一同纳入水资源可利用量考量后，布迪科框架在灌区中表现良好。在干旱与半干旱地区，灌溉水量每增加10%，蒸散发量可提升1.58%至5.29%；降水量每增加10%，蒸散发量可提升0.97%至4.25%。而在湿润与半湿润地区，蒸散发量的变化主要由能量供给驱动，潜在蒸发量每增加10%，蒸散发量可提升6.79%至8.61%。 基于上述8年观测数据，通过最小二乘法求得布迪科参数ω的最优取值。湿润与半湿润地区的ω参数值普遍大于干旱与半干旱地区。研究构建了经验方程，用以表征ω参数与NDVI及土壤性质之间的显著相关性。该方程所需输入因子简便易得，在复现ω参数取值时表现优异。