Water footprint implication of proposed future production scenarios - development of water availability indexes and issues in green water estimate

While water footprint of biofuel feedstocks has been quantified in a large body of studies, few have assessed its impact on water availabilities, especially green water (rainfall) availability, for potential large-scale... [ view full abstract ]

While water footprint of biofuel feedstocks has been quantified in a large body of studies, few have assessed its impact on water availabilities, especially green water (rainfall) availability, for potential large-scale biomass production. Given that plant growth critically depends on green water accessibility, it is important to have a generic water metric that evaluates whether green water supply may be a constraint for future biomass production in a certain area. However, few existing water availability metrics are dedicated to green water analysis. Here we propose a biophysical based green water availability indicator that is extended from existing indicators. A brief overview of water scarcity or availability indicators is also included. The proposed metric measures the fraction of rainfall consumptively used by potential biofuel feedstock within an area (e.g. watershed or county). Green water footprint (i.e. soil moisture losses caused by plant evapotranspiration (ET)) is usually estimated from effective rainfall (i.e. fraction of rainfall that is available to plant ET). Effective rainfall can be estimated using multiple methods, but many water footprint studies selected a certain method without evaluating uncertainties in resulted effective rainfall and therefore green water footprint estimation. To assess these uncertainties, we calculated county-level average monthly and annually effective rainfall for the conterminous United States using three different datasets and methods: (1) a water balance approach based on the NHDPlus V2 dataset, (2) the Smith’s method and (3) the USDA-SCS method. Preliminary results show that the three methods generate distinct spatial distribution patterns of estimated annual effective rainfall. Also, differences among the three sets of estimations are not evenly distributed throughout the year but are varying substantially from season to season. This study not only proposes an easily interpretable green water availability index, but it also highlights the importance of taking uncertainties in crop ET and effective rainfall estimation into account in green water availability assessments.