A modified water demand estimation method for drought identification over arid and semiarid regions

Reliable drought monitoring and analysis requires appropriate water demand estimates, particularly in arid and semiarid regions. Although potential evapotranspiration (PET) is commonly used to estimate water demand when monitoring droughts, the PET only considers the maximum atmospheric demand for water and ignores other variables that can affect surface water consumption and thus drought conditions. As such, the concept of climatically appropriate precipitation for existing conditions ((P) over cap), which considers soil moisture loss and recharge, runoff, and evapotranspiration, is developed as a modified water demand estimate in drought modeling. This study compared and evaluated the performance of PET and (P) over cap as water demand estimates in drought monitoring over arid and semiarid regions. The results show that the magnitudes of PET are much higher than precipitation (P) in the study region, particularly for the area with annual average P less than 300 mm, indicating that PET as a water demand indicator would lead to water demand being given too much weight in drought quantification. Accordingly, using the differences between P and PET to characterize surface water deficit/surplus could result in a water imbalance and inaccurate drought quantification in water-limited regions. In the physical world, water demand and supply are generally balanced over the long term, which is reasonably reflected by the differences between P and (P) over cap. The magnitudes and patterns of P and (P) over cap were generally similar to each other over the long term, meaning (P) over cap as water demand is physically more reasonable and reliable than PET in monitoring droughts. Moreover, the drought identification results based on the (P) over cap are more consistent with changes in observed streamflow and CPC soil moisture than the PET. These demonstrated that the (P) over cap as a modified water demand metric provides improved information for drought modeling in arid and semiarid regions. (C) 2015 Elsevier B.V. All rights reserved.