Terrestrial water storage deficits during the GRACE and GRACE-FO gap

This datasets provide predictions of the terrestrial water storage anomalies (TWSAs) during the 11-month gap (July 2017-May 2018) between the GRACE satellite and its follow-on GRACE-FO. The predictions were obtained by a hydroclimatic data-driven Bayesian convolutional neural network driven. The training period is April 2002-March 2014. Based on the predicted TWSAs, the terrestrial water storage deficit (WSD) dataset and its standardized version, the WSD index (WSDI) dataset, are also provided. The GIF animation depicts the GRACE and BCNN TWSAs in the 70 testing months (April 2014-June 2017 and June 2018-December 2020) and the 11-month gap (July 2017-May 2018).本数据集提供了GRACE卫星与其后续任务GRACE-FO之间11个月空白期（2017年7月至2018年5月）内的陆地水储量异常（terrestrial water storage anomalies，TWSAs）预测结果。该预测结果由基于水文气候数据驱动的贝叶斯卷积神经网络（Bayesian Convolutional Neural Network，BCNN）生成，模型训练时段为2002年4月至2014年3月。基于上述预测得到的陆地水储量异常数据，本数据集同步提供陆地水储量亏缺（terrestrial water storage deficit，WSD）数据集，及其标准化形式——陆地水储量亏缺指数（WSD Index，WSDI）数据集。本数据集附带的GIF动画展示了70个测试月（2014年4月至2017年6月、2018年6月至2020年12月）以及上述11个月空白期内GRACE与BCNN的陆地水储量异常结果。