Three drought indices (the Standardized Precipitation Index [SPI], Evaporative Stress Index [ESI], and Soil Moisture Anomaly Index [SMAI]) were integrated using triple collocation (TC) to produce the merged drought index (MDI). The new index was then compared with the Gravity recovery and climate experiment (GRACE)–Drought severity index (DSI), a comprehensive drought index reflecting storage variation in surface, sub-surface, and groundwater levels across East Asia and Australia, from 2003 to 2014. Before merging the three drought indices, their performance was analyzed. The mean correlation between the three drought indices and the GRACE–DSI indicated that the performance of the ESI was superior to the SMAI and SPI over the study areas. In terms of average weight results using the merging approach, the ESI was associated with larger weights (0.372 and 0.359) and contributions (43% and 38%), followed by the SMAI and SPI for East Asia and Australia, respectively. The SMAI achieved a similar weight (0.360) and contribution (39%) as the ESI across Australia. To determine the robustness of the MDI as estimated by TC weights, we evaluated the MDI and the reference GRACE-DSI with respect to documented drought records in the study areas. The MDI produced trends similar to those of the GRACE-DSI in Australia, while MDI and GRACE-DSI trends were not similar in East Asia. The correlation between the MDI and GRACE-DSI in Australia (0.41–0.62) was also higher than in East Asia (0.24–0.32) during the study periods. This discrepancy was due to the conceptual difference in that MDI reflects the near-surface water storage variation while GRACE-DSI reflects the variation of deeper water. Nevertheless, our results showed that the MDI out-performed single drought indices and was able to capture documented drought events across the study regions. This suggests that merging different drought indices into a single tool can better represent droughts, and may be a valuable approach for water resource management.

三个干旱指数（标准化降水指数 [SPI]、蒸发应力指数 [ESI] 和土壤水分异常指数 [SMAI]）使用三重搭配 (TC) 进行整合以产生合并干旱指数 (MDI)。然后将新指数与重力恢复和气候实验 (GRACE)-干旱严重程度指数 (DSI) 进行比较，这是一个综合干旱指数，反映了东亚和澳大利亚从 2003 年到2014. 在合并三个干旱指数之前，分析了它们的表现。三个干旱指数与 GRACE-DSI 之间的平均相关性表明，ESI 的表现优于研究区的 SMAI 和 SPI。就使用合并方法的平均权重结果而言，ESI 与更大的权重 (0. 372 和 0.359）和贡献（43% 和 38%），其次是东亚和澳大利亚的 SMAI 和 SPI。SMAI 的权重 (0.360) 和贡献 (39%) 与澳大利亚的 ESI 相似。为了确定由 TC 权重估计的 MDI 的稳健性，我们评估了 MDI 和参考 GRACE-DSI 相对于研究区域中记录的干旱记录。MDI 产生的趋势与澳大利亚的 GRACE-DSI 趋势相似，而东亚的 MDI 和 GRACE-DSI 趋势并不相似。在研究期间，澳大利亚的 MDI 和 GRACE-DSI 之间的相关性（0.41-0.62）也高于东亚（0.24-0.32）。这种差异是由于 MDI 反映近地表水储存变化而 GRACE-DSI 反映深水变化的概念差异。尽管如此，我们的结果表明，MDI 的表现优于单一干旱指数，并且能够捕获整个研究区域的记录干旱事件。这表明将不同的干旱指数合并为一个工具可以更好地代表干旱，并且可能是水资源管理的一种有价值的方法。