Application of the moisture balance drought index in the Colorado River Basin, USA

The Moisture Balance Drought Index (MBDI) was recently developed to monitor drought within the Colorado River Basin (CRB) of the southwestern United States. The MBDI is based on the balance of precipitation with the climatic demand for moisture, or potential evapotranspiration. The aim of the index is beyond representation of only precipitation, which can be misleading in warm climates or seasons, but short of representing soil moisture, which requires location-specific soil and vegetation characteristics. As such, the MBDI is capable of reflecting the impact of air temperature, and it can be used to directly compare conditions for different seasons and regions. Index values represent the cumulative difference of precipitation minus potential evapotranspiration calculated for various timeframes that relate to different impacts. The MBDI is expressed in percentiles for ease of interpretation in a historical perspective.

To validate the MBDI and to determine the value of representing climatic demand in drought monitoring, the index was compared with the precipitation-only Standardized Precipitation Index (SPI). First, the indices were matched with water supply data from select sub-basins of the CRB. Time series of watershed runoff, reservoir storage, and groundwater level were evaluated against time series of the drought indices. The SPI generally explains a greater percentage of the variance in runoff for the full historical record, but when focusing only on the dry half of the record the MBDI explains more of the variance for many of the basins studied. For reservoir storage, the SPI explains more of the historical variability for the larger capacity system of the two studied, especially during the cool season, but for the smaller capacity system the MBDI generally explains more of the variance, especially during the warm season. The MBDI explains a greater percentage of the variance in groundwater than the SPI at most of the well sites studied, although the two indices proved to be comparable. Second, the Normalized Difference Vegetation Index (NDVI) at select sites across Arizona was used to represent vegetation greenness, or health, which is most variable in spring across the region. Using the optimal 6-month timeframe for explaining variance in the spring NDVI, the MBDI and SPI are comparable, but when using the 3-month timeframe that defines the spring season, the MBDI explains significantly more variance. Slightly more of the variances in summer and fall greenness are explained by the SPI. Current work is focused on relating the indices to wildfire occurrences in the Southwest. The work to this point indicates that there is benefit to representing moisture balance in drought monitoring by including the temperature-driven climatic demand for water, especially given the recent and projected warming in many water-challenged regions.