A novel multivariate drought severity index: study of short-term hydrological signals within Amazon river basin

The pace of Earth’s climate warming obviously sped up, especially after 2000s. Droughts are increasingly becoming  frequent, longer and more severe, with lasting impacts on ecosystems, communities and people. Thus, addressing the problem of monitoring global (or regional) climate trends and  water storage changes is crucial. We propose a novel Multivariate Drought Severity Index (MDSI) estimated through the concept of Frank copulas that is based on DSIs determined from satellite-based geodetic data. The new multivariate approach is based on data provided by the Global Positioning System (GPS) and the Gravity Recovery and Climate Experiment (GRACE).

In this study, we analyze short-term (<9 months) signals of monthly-resampled vertical displacements for 25 GPS stations that are classified as benchmarks for hydrogeodesy within Amazon river basin. We show that despite GPS and/or GRACE limitations arising in data products or their quality, the GPS- and GRACE-based DSIs are characterized with a general coherent spatial pattern to the traditional climate indices (Standardized Precipitation Index (SPI) and Standardized Precipitation Evapotranspiration Index (SPEI)). Moreover, GPS- and GRACE-based DSIs are capable of capturing extreme hydrometeorological events reported for the Amazon basin. However, DSI variations from GPS and GRACE do not always reflect real hydrological changes as they could sometimes under- or overestimate them. Our analyses show that the newly proposed MDSI is a step towards strengthening  the credibility of combined GPS and GRACE data in drought assessment to improve  understanding of climate change impact on freshwater. We demonstrate that the MDSI recognizes the exact number of events, or one event less than index chosen as the most reliable for over 90% of selected stations. We notice that MDSI series are temporally consistent with extreme precipitation values. The wet and dry periods captured by MDSI are related with precipitation anomalies over 400 mm/month and below 100 mm/month, respectively.