Global agricultural green and blue water consumptive uses in the context of water scarcity and climate change

Agriculture, including rainfed and irrigated systems, is the single largest water user amongst all the economic sectors. There is an intrinsic linkage between water availability, food production and food security. This study takes a green and blue water perspective to model the water-food relations at the global level and on a high spatial resolution with a focus on Sub-Saharan Africa (SSA). The major issues investigated include: 1) green and blue water consumptive uses in crop production, particularly staple food crops at the global level; 2) possible impact of soil nutrient improvement and climate change on crop water uses and water productivity in SSA. A GIS-based EPIC model (GEPIC) is applied for the investigation for 26 major crops. GEPIC is a biophysical model that simulates plant growth and yield as a function of climate, soil, and crop management using a set of experimentally derived algorithms. The model estimates crop development on a daily time-step. Potential plant growth and yield are calculated first and subsequently multiplied by stress factors to obtain actual increases in biomass and yield. Besides plant development, nutrient cycling and changes on soil structure are simulated. The main functions of plant growth are light interception, conversion of energy and CO2 to biomass, and leaf area index (LAI) development. Growth is constrained by water, nutrient (N and P), temperature, salinity, and aeration stress. The 26 crops include cereals, legumes, oil crops, cotton, sugar cane and sugar beat, fruits and vegetables. The consumptive water use (CWU) in cropland is quantified in a spatially explicit way (30 arc-minute grid cell) by taking into account both green and blue water components. The results show that the global CWU was 5938 km(3) a(-1) in the crop growing periods and 7323 km(3) a(-1) in the entire year in cropland around the year 2000. Green water contributed to 84% of the global CWU in the crop growing periods and 87% of the global CWU on an annual basis. The high proportion of green water was due mainly to the dominance of rainfed agriculture, which consumed 4068 km(3) a(-1) of water in the crop growing periods and 5105 km(3) a(-1) of water in the entire year. In addition, in irrigated cropland, green water contributed to 50% of the total CWU in the crop growing periods, and over 60% of the annual total CWU. The consumptive blue water use (CBWU) was 927 km(3) a(-1) in cropland on a global scale based on land cover and climate data around the year 2000. In crop growing period, blue water accounts for 16% of the global CWU. High CBWU occurs in Northern and Southern India, Eastern part of China, and the Mid Central of the USA. These regions are the major agricultural production regions in the world, and they also have very high CWU. As for the blue water proportion, regions with high values are located in the northern part of China, several West Asian countries, Middle East and North Africa, the western part of the USA, and Chile. These regions mostly have arid or semi-arid climate with low precipitation, which can only meet part of the water required by crops. In order to achieve high crop yields, irrigation water has to be supplied in addition to precipitation. Largely due to the low precipitation, irrigation depth is generally very high, resulting in high blue water proportion in these regions. SSA is currently the region with the lowest yields and the most serious problems in food insecurity. Poor water management and soil nutrient depletion have been partly the causes. Climate change is expected to worsen the situation if no action is taken. The GEPIC model is applied to investigate impact of nutrient improvement on consumptive water uses under the current and future climate conditions. The results show that the crop water productivity (and yields) will increase significantly with sufficient nutrient input. This can be achieved with little increase in consumptive water use in most of the regions (holding the land use pattern constant) as the result of the vapor shift from evaporation to transpiration in the total evapotranspiration by crops.