Statistical assessment of temporal aliasing in GRACE/-FO and NGGM induced by hydrometeorological extremes

Abstract

Hydrometeorological extreme events, such as heavy rainfall, can cause a rapid increase in local water storage over a short period of time. The duration, magnitude, and area of the event, as well as the runoff of the affected region, determine the life span of these water mass changes. This leads to temporal aliasing, which, along with instrument noise, poses a significant challenge to improving the accuracy and resolution of satellite gravimetry products from GRACE, GRACE Follow-On, and next-generation gravity missions (NGGM). The current Atmospheric Ocean Dealiasing (AOD) products can remove tidal and sub-monthly non-tidal mass variations in the ocean and atmosphere from the level-1 GRACE/-FO data. However, the reanalysis and forecast datasets underlying these AOD products only have a limited temporal and spatial resolution and do not include high-frequency hydrological mass variations and liquid cloud water content of atmospheric mass, that can significantly increase during hydrometeorological extreme events. The research group New Refined Observations of Climate Change from Spaceborne Gravity Missions (NEROGRAV), funded by the German Research Foundation (DFG), aims at improving GRACE/-FO data products by developing new analysis methods and modeling approaches. This includes a revision of existing geophysical background models as well as their spatial-temporal parameterization. Within this research group we investigate how hydrometeorological extreme events are mapping into GRACE/-FO level-1 data. In this study, we provide statistics on events that may affect GRACE/-FO/NGGM observations and are not accounted for in current AOD products. Using a 3D connected component algorithm, we determine the duration, magnitude, and area of these events for multiple test years over the entire GRACE/-FO observation time span from 2002 to 2023. As expected, the majority of hydrometeorological extreme events take place in tropical regions and areas that are frequently impacted by typhoons, hurricanes, and monsoon rains such as Japan, northern India, and around the Gulf of Mexico. Additionally, we found an increasing number of events over Europe that could significantly impact GRACE/-FO/NGGM observations. Overall, the number of events per year more than doubles from the start to the end of the observation period, which is likely due to climate change. We suspect that the GRACE-FO LRI measurement is sensitive to instantaneous precipitation and liquid cloud water content during overfly for a significant number of these extreme events. Hence, we anticipate an increased probability of NGGM observations being affected by such events, particularly in the context of ongoing climate change. Therefore, we recommend that upcoming dealiasing procedures carefully consider these events.