Global-scale compound Droughts and heatwaves: inland/coastal type grouping, diversity of temperature extremes, and dynamically-based Interpretable reconstruction

Although compound drought and heatwave extremes have recently drawn much attention globally, there exist three interesting issues (i.e., event detection, temperature diversity, and interpretable reconstruction) to explore as follows: --First, as drought events can spread over space and evolve over time, how can we perform event detection as accurately as possible? Are there differences in coastal/inland regions?  --Second, whether droughts are always concurrent with heatwaves remains unknown. Moreover, how temperature abnormalities evolve spatiotemporally during drought development and how their associated categories are distributed globally are not fully understood. --Third, it is common sense that droughts and associated near-surface temperature anomalies can be attributed to amplified vertical subsidence and anomalous anticyclonic circulations from dynamic perspectives. However, one open and interesting issues remain unknown: That is, whether hydrometeorological situations under droughts can be reproduced directly utilizing variability of atmospheric dynamics and what specific roles atmospheric dynamics play in drought reconstruction.

 

To explore the three issues mentioned above, our recent achievements are as follows:

-- First, regarding accurate event detection and type division, we identified global-scale seasonal-scale meteorological drought events following the recently proposed 3D DBSCAN-based workflow of event detection. The 3D DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm can directly obtain arbitrarily shaped point collections over a given 3D space. Subsequently, these detected drought events are further grouped into inland and coastal types, as the observations revealed that some droughts over coastal regions originate from, extend to, or are accompanied by long-term precipitation deficits over adjacent oceans. [see algorithm cases (https://doi.org/10.1016/j.aosl.2022.100324), Glo3DHydroClimEventSet(v1.0) products (https://doi.org/10.1002/joc.8289) , and global drought detection (https://spj.science.org/doi/10.34133/olar.0016 ) ]

--Second, regarding diversity of temperature extremes compounded with droughts, we investigated this fundamental issue from the perspectives of temperature abnormality–based drought classification and statistical characteristics of process evolution. The major procedures and achievements were as follows. First, the detected global-scale 3D DBSCAN-based drought events of our study were employed and assigned to Hot, Cold, Normal, and Hybrid categories utilizing a self-designed temperature abnormality–based classification algorithm; the associated global-scale occurrences of these four event categories were approximately 40%, 10%, 30%, and 20%, respectively, and in turn, they displayed statistically significant (p value < 0.05) increasing, decreasing, decreasing, and increasing trends, respectively, during 1980–2020.   [see diversity of temperature anomalies (https://spj.science.org/doi/10.34133/olar.0017 ) ]

--Third, regarding dynamically-based reconstruction of compound droughts and heatwaves, we employs three kinds of dynamic features (i.e., vertical velocity, relative vorticity, and horizontal divergence) for hydrometeorological reconstruction (e.g., precipitation and near-surface air temperature) under drought situations through a so-called XGBoost (extreme gradient boosting) ensemble learning method. The study adopts the reconstruction scheme on the interannual variability and finds dynamically based reconstruction feasible, seemingly regardless of seasonality and drought-inducing mechanisms. More importantly, from interpretable perspectives, global-scale analysis of dynamic contributions helps discover unexpected dynamic drought-inducing roles and associated latitudinal modulation. That is, low-level cyclonic/anticyclonic anomalies contribute to drought development in the northern middle and high latitudes, while upper-level vertical subsidence contributes significantly to tropical near-surface temperature anomalies concurrent with droughts. [see paper (https://doi.org/10.1175/JHM-D-22-0006.1)]