Global Precipitation Nowcasting of Integrated Multi-satellitE Retrievals for GPM: A U-Net Convolutional LSTM Architecture

Reyhaneh Rahimi aDepartment of Civil Environmental and Geo-Engineering and the Saint Anthony Falls Laboratory, University of Minnesota

Search for other papers by Reyhaneh Rahimi in
Current site
Google Scholar
PubMed
Close
,
Praveen Ravirathinam bDepartment of Computer Science and Engineering, University of Minnesota

Search for other papers by Praveen Ravirathinam in
Current site
Google Scholar
PubMed
Close
,
Ardeshir Ebtehaj aDepartment of Civil Environmental and Geo-Engineering and the Saint Anthony Falls Laboratory, University of Minnesota

Search for other papers by Ardeshir Ebtehaj in
Current site
Google Scholar
PubMed
Close
,
Ali Behrangi cDepartment of Hydrology and Atmospheric Sciences, The University of Arizona

Search for other papers by Ali Behrangi in
Current site
Google Scholar
PubMed
Close
,
Jackson Tan dEarth Sciences Division, NASA’s Goddard Space Flight

Search for other papers by Jackson Tan in
Current site
Google Scholar
PubMed
Close
, and
Vipin Kumar bDepartment of Computer Science and Engineering, University of Minnesota

Search for other papers by Vipin Kumar in
Current site
Google Scholar
PubMed
Close
Restricted access

Abstract

This paper presents a deep supervised learning architecture for 30 min global precipitation nowcasts with a 4-hour lead time. The architecture follows a U-Net structure with convolutional long short-term memory (ConvLSTM) cells empowered by ConvLSTM-based skip connections to reduce information loss due to the pooling operation. The training uses data from the Integrated MultisatellitE Retrievals for GPM (IMERG) and a few key drivers of precipitation from the Global Forecast System (GFS). The impacts of different training loss functions, including the mean-squared error (regression) and the focal-loss (classification), on the quality of precipitation nowcasts are studied. The results indicate that the regression network performs well in capturing light precipitation (<1.6 mm hr−1) while the classification network can outperform the regression counterpart for nowcasting of high-intensity precipitation (>8 mm hr−1), in terms of the critical success index (CSI). It is uncovered that including the forecast variables can improve precipitation nowcasting, especially at longer lead times in both networks. Taking IMERG as a relative reference, a multi-scale analysis, in terms of fractions skill score (FSS), shows that the nowcasting machine remains skillful for precipitation rate above 1 mm hr−1 at the resolution of 10 km compared to 50 km for GFS. For precipitation rates greater than 4 mm hr−1, only the classification network remains FSS-skillful on scales greater than 50 km within a 2-hour lead time.

© 2024 American Meteorological Society. This is an Author Accepted Manuscript distributed under the terms of the default AMS reuse license. For information regarding reuse and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Ardeshir Ebtehaj, email: ebtehaj@umn.edu

Abstract

This paper presents a deep supervised learning architecture for 30 min global precipitation nowcasts with a 4-hour lead time. The architecture follows a U-Net structure with convolutional long short-term memory (ConvLSTM) cells empowered by ConvLSTM-based skip connections to reduce information loss due to the pooling operation. The training uses data from the Integrated MultisatellitE Retrievals for GPM (IMERG) and a few key drivers of precipitation from the Global Forecast System (GFS). The impacts of different training loss functions, including the mean-squared error (regression) and the focal-loss (classification), on the quality of precipitation nowcasts are studied. The results indicate that the regression network performs well in capturing light precipitation (<1.6 mm hr−1) while the classification network can outperform the regression counterpart for nowcasting of high-intensity precipitation (>8 mm hr−1), in terms of the critical success index (CSI). It is uncovered that including the forecast variables can improve precipitation nowcasting, especially at longer lead times in both networks. Taking IMERG as a relative reference, a multi-scale analysis, in terms of fractions skill score (FSS), shows that the nowcasting machine remains skillful for precipitation rate above 1 mm hr−1 at the resolution of 10 km compared to 50 km for GFS. For precipitation rates greater than 4 mm hr−1, only the classification network remains FSS-skillful on scales greater than 50 km within a 2-hour lead time.

© 2024 American Meteorological Society. This is an Author Accepted Manuscript distributed under the terms of the default AMS reuse license. For information regarding reuse and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Ardeshir Ebtehaj, email: ebtehaj@umn.edu
Save