The Standardized Vertical Velocity Anomaly Index (SVVAI): Using Atmospheric Dynamical Anomalies to Simulate and Predict Meteorological Droughts
Abstract. Vertically downward motion of air current is physically drought-inducing, which has the potential of being a simple and universal drought indicator. The core objective of the present study is to employ vertical motion to simulate and predict droughts after investigating dynamically drought-inducing mechanism. Season-scale drought processes and spatial distributions during 2009–2016 are our concerns, and all the drought study regions of China were chosen as the research areas. Three-month SPI (SPI3) updated daily was used to identify drought processes, and original vertical motion and associated horizontal divergence were also transformed to season-scale standardized anomalies (SA) with a daily running window. In situ observation, ERA-Interim reanalysis, and CFSv2 forecast products were comprehensively employed for drought simulation and prediction. To date, the main results and conclusions are as follow: (1) Atmospheric dynamical anomalies during drought processes and key phases were uncovered. Dynamically drought-inducing features are generally characterized as the typically anomalous
upper-convergence–lower-divergence patterns and the intensified downward vertical motion as expected. Signal intensities and vertical configurations are time-varying and seemingly coincide with evolution of regional processes. Particularly, vertical velocity exhibited universally strengthened downward anomalies over almost all the droughts. (2) On the basis of dynamically vertical features uncovered above, the SVVAI (Standardized Vertical Velocity Anomaly Index) is newly proposed. The SVVAI is calculated using SA-based values of vertical motion at multiple pressure levels in the troposphere. The SVVAI_ave and SVVAI_max, corresponding to the vertically average- and maximum-based computation schemes, can be adopted. (3) Drought processes and spatial distributions were simulated with the SVVAI_ave and SVVAI_max. They commonly show highly positive correlations with realistic ones over most regions, and the SVVAI_ave outperformed the SVVAI_max. (4) To further understand difference of simulation capacity, temporal correlation coefficients (TCC) of the SVVAI_ave against observed SPI3 at the grid scale were used for analysis. Positive TCCs above +0.3 occupies most areas to the east of 110° E, while large-area low TCCs (−0.1 ~ +0.3) appear to the west of 110° E over China. It is notably seen that East China and Northeast China are the two regions with highly positive TCCs (+0.6 ~ +0.8). (5) Drought prediction using the SVVAI_ave was preliminarily explored. Regarding the prospective 60-day process prediction, the predicted SVVAI_ave was equally matched with or a little better than the forecasted SPI3 in most cases. Predicted spatial distribution is preliminarily assessed via the example of the 2011 summer–autumn drought over Southwest China, and prediction performance at the occurrence, peak and termination times are inconsistent. (6) Overall, the novel SVVAI herein may be complementary to current approaches of operational drought monitoring and prediction. Further study could be focused on the two following aspects: One is index applicability, that is to say, to explore when and where the predicted SVVAI outperforms the forecasted SPI3. The other is to further explore antecedent drought-inducing signals of atmospheric/oceanic anomalies with the bridge of vertical motion, which may provide a fundamental approach for drought prediction with long lead times.
Viewed (geographical distribution)