Bias correction of GPM IMERG Early Run daily precipitation product using near real-time CPC global measurements
(2022) In Atmospheric Research 279.- Abstract
This study focused on improving the performance of the near real-time Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) Early Run (IMERG-E) product based on a newly developed bias-correction scheme, LSCDF. The LSCDF was established by integrating the mean-based Linear Scaling (LS) and quantile-mapping Cumulative Distribution Function (CDF) matching approaches. The daily updating gauge-based precipitation data from the Climate Prediction Center (CPC) unified analysis were used as the benchmark for bias correction. The IMERG-E bias-corrected precipitation data were evaluated against the daily ground measurements from 807 meteorological stations across mainland China and compared with the raw IMERG-E and... (More)
This study focused on improving the performance of the near real-time Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) Early Run (IMERG-E) product based on a newly developed bias-correction scheme, LSCDF. The LSCDF was established by integrating the mean-based Linear Scaling (LS) and quantile-mapping Cumulative Distribution Function (CDF) matching approaches. The daily updating gauge-based precipitation data from the Climate Prediction Center (CPC) unified analysis were used as the benchmark for bias correction. The IMERG-E bias-corrected precipitation data were evaluated against the daily ground measurements from 807 meteorological stations across mainland China and compared with the raw IMERG-E and IMERG Final Run (IMERG-F; research-level with gauge calibration) retrievals. Evaluation results for the period 2015 to 2017 showed that the LSCDF method effectively improves near real-time IMERG-E precipitation estimates at the daily scale. The bias-corrected IMERG-E precipitation estimates were in significantly better agreement with ground measurements than the original IMERG-E at the point-daily resolutions, with the correlation coefficient increasing from 0.66 to 0.77, relative bias decreasing from 11.1% to −3.1%, and root-mean-square error dropping from 6 mm/day to 4.39 mm/day. In addition, the bias-corrected IMERG-E was apparently superior to IMERG-E in detecting precipitation events at various intensity levels including small, light, moderate, and heavy precipitation. Although IMERG-E tended to significantly underestimate or overestimate precipitation during three typical typhoons, the bias-corrected IMERG-E can match the rainfall spatial distributions well, demonstrating a considerable capability in capturing the extreme precipitation. Generally, the bias-corrected IMERG-E featured a substantial improvement over the original IMERG-E, and it even exhibited a more satisfactory overall performance than the research-level gauge-calibrated IMERG-F product. Our study demonstrates that the bias-correction method LSCDF can improve the accuracy of satellite precipitation products to benefit the near real-time application of satellite products for natural hazards.
(Less)
- author
- Wei, Linyong ; Jiang, Shanhu ; Ren, Liliang ; Zhang, Linqi ; Wang, Menghao ; Liu, Yi and Duan, Zheng LU
- organization
- publishing date
- 2022-12-01
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Bias correction, CPC, IMERG, LSCDF, Satellite precipitation
- in
- Atmospheric Research
- volume
- 279
- article number
- 106403
- publisher
- Elsevier
- external identifiers
-
- scopus:85136604642
- ISSN
- 0169-8095
- DOI
- 10.1016/j.atmosres.2022.106403
- language
- English
- LU publication?
- yes
- id
- 21391a58-ac0c-45f4-8c41-be9dd21b1c41
- date added to LUP
- 2022-10-17 08:50:53
- date last changed
- 2023-05-10 11:06:45
@article{21391a58-ac0c-45f4-8c41-be9dd21b1c41,
abstract = {{<p>This study focused on improving the performance of the near real-time Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) Early Run (IMERG-E) product based on a newly developed bias-correction scheme, LSCDF. The LSCDF was established by integrating the mean-based Linear Scaling (LS) and quantile-mapping Cumulative Distribution Function (CDF) matching approaches. The daily updating gauge-based precipitation data from the Climate Prediction Center (CPC) unified analysis were used as the benchmark for bias correction. The IMERG-E bias-corrected precipitation data were evaluated against the daily ground measurements from 807 meteorological stations across mainland China and compared with the raw IMERG-E and IMERG Final Run (IMERG-F; research-level with gauge calibration) retrievals. Evaluation results for the period 2015 to 2017 showed that the LSCDF method effectively improves near real-time IMERG-E precipitation estimates at the daily scale. The bias-corrected IMERG-E precipitation estimates were in significantly better agreement with ground measurements than the original IMERG-E at the point-daily resolutions, with the correlation coefficient increasing from 0.66 to 0.77, relative bias decreasing from 11.1% to −3.1%, and root-mean-square error dropping from 6 mm/day to 4.39 mm/day. In addition, the bias-corrected IMERG-E was apparently superior to IMERG-E in detecting precipitation events at various intensity levels including small, light, moderate, and heavy precipitation. Although IMERG-E tended to significantly underestimate or overestimate precipitation during three typical typhoons, the bias-corrected IMERG-E can match the rainfall spatial distributions well, demonstrating a considerable capability in capturing the extreme precipitation. Generally, the bias-corrected IMERG-E featured a substantial improvement over the original IMERG-E, and it even exhibited a more satisfactory overall performance than the research-level gauge-calibrated IMERG-F product. Our study demonstrates that the bias-correction method LSCDF can improve the accuracy of satellite precipitation products to benefit the near real-time application of satellite products for natural hazards.</p>}},
author = {{Wei, Linyong and Jiang, Shanhu and Ren, Liliang and Zhang, Linqi and Wang, Menghao and Liu, Yi and Duan, Zheng}},
issn = {{0169-8095}},
keywords = {{Bias correction; CPC; IMERG; LSCDF; Satellite precipitation}},
language = {{eng}},
month = {{12}},
publisher = {{Elsevier}},
series = {{Atmospheric Research}},
title = {{Bias correction of GPM IMERG Early Run daily precipitation product using near real-time CPC global measurements}},
url = {{http://dx.doi.org/10.1016/j.atmosres.2022.106403}},
doi = {{10.1016/j.atmosres.2022.106403}},
volume = {{279}},
year = {{2022}},
}