Developing and Exploiting a New Global Reanalysis of Evaporative Demand for Global Food-Security Assessments and Drought Monitoring.

In drought monitoring, the supply of moisture to the surface has been well represented by precipitation, for which data are widely available; however, until recently, accurate representation of the demand side of drought has suffered from the paucity of driver data required to fully characterize evaporative demand (E₀; the “thirst of the atmosphere”). Instead, E₀ has often been replaced by its climatological mean, estimated by non-physically based models, or has been neglected entirely. This deficient modeling has not only obstructed the exploitation of E₀’s early warning and monitoring potential over such regions as North America, but it is particularly troublesome over data-sparse regions that are also home to drought-vulnerable and food-insecure populations, such as across much of Africa. There is thus urgent need for global E₀ estimates for physically accurate drought analyses and food security assessments; further we need an improved understanding of how E₀ and drought interact and to exploit these interactions in drought monitoring.

This presentation will summarize NOAA’s current efforts to meet these needs. NOAA is working in partnership with the Famine Early Warning System Network (FEWS NET) and the Climate Hazards Group at the University of California Santa Barbara (UCSB) and regional scientists stationed in food-insecure countries. Exploiting the advent of accurate, fine-resolution land-surface/atmosphere global reanalyses such as MERRA-2, we have developed a >38-year, daily, global Penman-Monteith reference evapotranspiration dataset as a fully physical metric of E₀. For drivers we use temperature, humidity, solar radiation, and wind speed drivers from MERRA-2. We have also developed a novel drought index based on this dataset, the Evaporative Demand Drought Index (EDDI) to represents drought’s demand perspective. This reference ET reanalysis and the EDDI product are gaining traction within the food-security assessment community, primarily as an input to agricultural assessments and in providing early warning and ongoing monitoring of agricultural flash drought and hydrologic drought. The significance of these sorts of assessments is that the probabilistic approach projects drought risk in a novel manner--it attributes probability to the occurrence and impacts of future droughts of different severity--and spatial aggregation of such risk communicates more clearly to stakeholders. Beyond these uses, we are also collaborating across borders within North America, specifically to wildfire and agricultural stakeholders from Canada for the North American Drought Monitor.

I will highlight the potential of this new, global reference evapotranspiration reanalysis in under-gaged areas and as an input to various drought-monitoring and food-security products, including the use of EDDI as a global drought early warning tool. Specifically, I will summarize the development of the reference evapotranspiration reanalysis, including its spatial downscaling to a high resolution; the challenges faced in, and early results of, its verification across Africa; early results of its use in food security assessments; and the development and potential of the EDDI product.