WRF Nudging, Diffusion Probabilistic Postprocessing, and Distributional Reliability Analysis: A Case Study of the 3 July 2024 Extreme Precipitation Event in Lagos–Ogun, Nigeria

Extreme precipitation over rapidly urbanizing West African coastal regions remains difficult to simulate because rainfall is influenced by multiscale atmospheric forcing, land conditions, and convective processes. This study examines the 3 July 2024 extreme precipitation event in Lagos–Ogun, Nigeria, using high-resolution Weather Research and Forecasting (WRF) simulations, satellite observations, conditional diffusion probabilistic postprocessing, and distributional reliability analysis. Three WRF experiments were evaluated against Integrated Multi-satellitE Retrievals for the Global Precipitation Measurement mission (IMERG): (1) an ERA5-forced control with ERA5-derived initial soil moisture and no nudging; (2) a grid-nudged experiment retaining the same soil-moisture initialization, with nudging on the 27 and 9 km domains but not the 3 km domain; and (3) a non-nudged experiment with SMAP-informed initial soil moisture. Relative to the ERA5 baseline, grid nudging produced the largest physical improvement, reducing event-scale root mean square error from 55.70 to 28.23 mm and rainfall-centroid displacement from 61.67 to 32.06 km. SMAP-informed soil-moisture initialization produced substantially smaller improvements. Conditional diffusion substantially reduced residual prediction error, achieving an out-of-fold hourly root mean square error of 2.80 mm, continuous ranked probability score of 0.894 mm, and raw 90% interval coverage of 88.70%. Calibration increased 90% interval coverage to 91.50% overall and 92.38% during the primary event period. Distributional Reliability Explanation Attribution (DREA) identified the 10 m wind vector as the dominant and most stable influence on the postprocessed predictive distribution across 80 model realizations. The agreement between the nudging experiment and DREA indicates that the simulated rainfall and fitted predictive distribution were more sensitive to atmospheric conditions than to the tested soil-moisture modification during this event, without establishing physical causality. These findings demonstrate the potential of the integrated workflow for this case, while its predictive robustness and transferability to other storms and regions remain untested.

Authors

Institutions

Publication Details

Journal
Atmosphere
Published
2026-11-18
DOI
https://doi.org/10.3390/atmos17111139
Primary Topic
Meteorological Phenomena and Simulations
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

WRF Nudging, Diffusion Probabilistic Postprocessing, and Distributional Reliability Analysis: A Case Study of the 3 July 2024 Extreme Precipitation Event in Lagos–Ogun, Nigeria

Chibuike Chiedozie Ibebuchi, Richard Boateng Bonsu
2 links
Atmosphere
Meteorological Phenomena and Simulations
article

WRF Nudging, Diffusion Probabilistic Postprocessing, and Distributional Reliability Analysis: A Case Study of the 3 July 2024 Extreme Precipitation Event in Lagos–Ogun, Nigeria

Chibuike Chiedozie Ibebuchi, Richard Boateng Bonsu
article en

Abstract

Extreme precipitation over rapidly urbanizing West African coastal regions remains difficult to simulate because rainfall is influenced by multiscale atmospheric forcing, land conditions, and convective processes. This study examines the 3 July 2024 extreme precipitation event in Lagos–Ogun, Nigeria, using high-resolution Weather Research and Forecasting (WRF) simulations, satellite observations, conditional diffusion probabilistic postprocessing, and distributional reliability analysis. Three WRF experiments were evaluated against Integrated Multi-satellitE Retrievals for the Global Precipitation Measurement mission (IMERG): (1) an ERA5-forced control with ERA5-derived initial soil moisture and no nudging; (2) a grid-nudged experiment retaining the same soil-moisture initialization, with nudging on the 27 and 9 km domains but not the 3 km domain; and (3) a non-nudged experiment with SMAP-informed initial soil moisture. Relative to the ERA5 baseline, grid nudging produced the largest physical improvement, reducing event-scale root mean square error from 55.70 to 28.23 mm and rainfall-centroid displacement from 61.67 to 32.06 km. SMAP-informed soil-moisture initialization produced substantially smaller improvements. Conditional diffusion substantially reduced residual prediction error, achieving an out-of-fold hourly root mean square error of 2.80 mm, continuous ranked probability score of 0.894 mm, and raw 90% interval coverage of 88.70%. Calibration increased 90% interval coverage to 91.50% overall and 92.38% during the primary event period. Distributional Reliability Explanation Attribution (DREA) identified the 10 m wind vector as the dominant and most stable influence on the postprocessed predictive distribution across 80 model realizations. The agreement between the nudging experiment and DREA indicates that the simulated rainfall and fitted predictive distribution were more sensitive to atmospheric conditions than to the tested soil-moisture modification during this event, without establishing physical causality. These findings demonstrate the potential of the integrated workflow for this case, while its predictive robustness and transferability to other storms and regions remain untested.

AtmosphereVol. 17(11)
Morgan State University (US)
Climate action, Clean water and sanitation
Meteorological Phenomena and Simulations
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon