Agricultural drought assessment in data-limited arid regions using opensource remotely sensed data: a case study from Jordan

Data availability is a key restriction in drought monitoring, with rainfall-based methods being more widely used despite their shortcomings, primarily due to the lack of observations for other meteorological parameters. This study evaluates open-source remote sensing based meteorological data from the POWER database for agricultural drought assessment in Jordan, where data access is limited, and drought-attributed crop failures are increasingly reported. The results showed strong correlations with ground measurements for precipitation (0.67-0.91) and temperature (> 0.99) from POWER. Standardized Precipitation and Standardized Precipitation Evapotranspiration indices (SPI and SPEI) calculated using POWER data showed acceptable mean absolute errors (0.3-0.9) based on gauged measurements from several semiarid sites. During periods of reported crop failures, precipitation was above historical values and SPI indicated wet conditions, while SPEI reflected extreme droughts that supported the reports of local farmers. Correlations between drought patterns produced using POWER and crop yields were significant (P < 0.5), this was further supported by pairwise t-tests which demonstrated a significant impact of drought on crop production (P = 0.017). These were consistent with previous agricultural drought assessments done using gauged measurements over drought vulnerable sites in the region. The results of this study suggest the need for revising current protocols relying solely on rainfall for drought assessment. The results also add to the existing body of literature supporting the adoption of SPEI for agricultural drought impact assessment and complementing gauged observations with open-source remote sensing data. Data from POWER seems to be suitable for this purpose under proper technical site-specific evaluations.

关键词