Evaluation of High-Resolution Precipitation Estimates from Satellites during July 2012 Beijing Flood Event Using Dense Rain Gauge Observations

Chen, Sheng; Huijuan, Liu; You, Yalei; Mullens, Esther; Hu, Junjun; Yuan, Ye; Huang, Mengyu; He, Ling‐Yun; Luo, Yongming; Xing-ji, Zeng; Tang, Guoqiang; Hong, Yang

doi:10.1371/journal.pone.0089681

Cited by 47 publications

(20 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…R4 presents a hit bias comparable to R6 (~1.2), but for other Koppen climatic regions it is slightly less. This corroborates with what shown in past literature, which suggests that IR‐dominant algorithms, such as PERSIANN‐CCS, exhibit a bias to extreme precipitation events (Chen et al, ).…”

Section: Resultssupporting

confidence: 92%

Investigating the Potential of Using Satellite‐Based Precipitation Radars as Reference for Evaluating Multisatellite Merged Products

2018

View full text Add to dashboard Cite

This work investigates the viability of using the Global Precipitation Mission (GPM) Dual‐frequency Precipitation Radar (DPR) as a reference for evaluating multisatellite precipitation products in locations where a ground‐based reference is not available. The Integrated Multi‐Satellite Retrievals for Global Precipitation Measurement (IMERG) version V05 products (early, late, and final) and the microwave‐only (MW) and infrared‐only (IR) components are evaluated against two reference data sets, derived from the Multi‐Radar/Multi‐Sensor System suite of products (MRMS) and the level‐2 DPR (2ADPR). The analysis focuses on a 2‐year period (2014–2015), excluding winters to evaluate liquid‐phase precipitation only. Systematic and random errors are assessed between the satellite‐based products and the MRMS‐based reference. Systematic error for the IMERG early, late, and final products and the MW component are shown to be comparable to those of 2ADPR, while the IR component exhibits a larger bias. Random errors of the IR estimates are found to be 1 order of magnitude larger than the 2ADPR random error. Thus, the hypothesis of using 2ADPR as a benchmark for quantifying uncertainties associated with IMERG holds true for evaluating random errors associated with IR precipitation estimates. Errors in the IR component are then investigated as a function of climatology and seasonality across contiguous United States. The temperate, oceanic, and subtropical climates show the best performance in terms of probability of detection and success ratio, whereas arid‐desert manifests the smallest root‐mean‐square errors. This work suggests that climatic zone‐specific error characterization model is necessary to estimate uncertainties associated with the IMERG products.

show abstract

Section: Resultssupporting

confidence: 92%

Investigating the Potential of Using Satellite‐Based Precipitation Radars as Reference for Evaluating Multisatellite Merged Products

2018

View full text Add to dashboard Cite

show abstract

“…Numerous previous studies have evaluated different satellite products over China [18,35,67,68] and several studies have pointed out differences between different versions of the same product [43,67]. However, none of these studies cover the different versions of GPM-based satellite precipitation products over China.…”

Section: Discussionmentioning

confidence: 99%

Comprehensive Evaluation of Two Successive V3 and V4 IMERG Final Run Precipitation Products over Mainland China

et al. 2017

View full text Add to dashboard Cite

Abstract:The Integrated Multi-satellitE Retrievals for Global Precipitation Measurement Final Run (IMERGF) product has now been upgraded to Version 4 (V4), which has been available since March 2017. Therefore, it is desirable to evaluate the characteristic differences between the V4 and the previous V3 products. A comprehensive performance evaluation of the errors of the successive V3 and V4 IMERGF products is performed with a comparison of the China daily Precipitation Analysis Products (CPAP) from March 2014 to February 2015. The version 6 Global Satellite Mapping of Precipitation (GSMaP) research product (which is another Global Precipitation Measurement (GPM) based precipitation product) is also used as a comparison in this study. Overall, the IMERGF-V4 product does not exhibit the anticipated improvement for China compared to the IMERGF-V3 product. An analysis of the metrics of annual daily average precipitation over China for the IMERGF-V3 and IMERGF-V4 products indicates a decrease of the relative bias (RB) from 3.70% to −7.18%, a decrease of the correlation coefficient (CC) from 0.91 to 0.89, an increase of the fractional standard error (FSE) from 0.49 to 0.56, and an increase of the root-mean-square error (RMSE) from 0.63 mm to 0.72 mm. Compared to the IMERGF-V3 product, the IMERGF-V4 product exhibits a significant underestimation of precipitation in the Qinghai-Tibetan plateau with a much lower RB of −60.91% (−58.19%, −65.30%, and −63.74%) based on the annual (summer, autumn, and winter) daily average precipitation and an even worse performance during winter (−72.33% of RB). In comparison, the GSMaP product outperforms the IMERGF-V3 and IMERGF-V4 products and has the smallest RMSE (0.47 mm/day), highest CC (0.95), lowest FSE (0.37), and best performance of the RB (−2.39%) in terms of annual daily precipitation over China. However, the GSMaP product underestimates the precipitation more than the IMERGF-V3 product for the arid XJ region.

show abstract

“…A number of global precipitation data sets do exist, based on satellite data [e.g., Huffman et al ., ; Joyce et al ., ], model reanalysis [e.g., Dee et al ., ] or gauge records [e.g., Xie et al ., ]. However, these products are known to have limitations that are of particular concern to flood modeling including spatially variable biases [ Kidd et al ., ], poor correlation with ground gauges at short (∼daily) time scales [ Chen et al ., ; Cohen Liechti et al ., ], poor representation of spatial variability over smaller catchments [ He et al ., ] and a tendency to underestimate heavy rainfall [ Chen et al ., ; Gao and Liu , ]. Sampson et al .…”

Section: Global Flood Hazard Modeling: Six Key Challengesmentioning

confidence: 99%

“…A number of global precipitation data sets do exist, based on satellite data [e.g., Huffman et al, 2007;Joyce et al, 2004], model reanalysis [e.g., Dee et al, 2011] or gauge records [e.g., Xie et al, 2007]. However, these products are known to have limitations that are of particular concern to flood modeling including spatially variable biases [Kidd et al, 2012], poor correlation with ground gauges at short (daily) time scales [Chen et al, 2014;Cohen Liechti et al, 2012], poor representation of spatial variability over smaller catchments [He et al, 2009] and a tendency to underestimate heavy rainfall [Chen et al, 2014;Gao and Liu, 2013]. Sampson et al [2014] evaluated the effect of these differences on flood risk using a cascade model structure that replicates an insurance catastrophe model and found estimates of monetary loss from flooding to vary by more than an order of magnitude depending on whether the cascade was driven with gauge, radar, satellite, or reanalysis data.…”

Section: Extreme Flow Generation 221 Reviewmentioning

confidence: 99%

A high-resolution global flood hazard model

et al. 2015

View full text Add to dashboard Cite

Floods are a natural hazard that affect communities worldwide, but to date the vast majority of flood hazard research and mapping has been undertaken by wealthy developed nations. As populations and economies have grown across the developing world, so too has demand from governments, businesses, and NGOs for modeled flood hazard data in these data‐scarce regions. We identify six key challenges faced when developing a flood hazard model that can be applied globally and present a framework methodology that leverages recent cross‐disciplinary advances to tackle each challenge. The model produces return period flood hazard maps at ∼90 m resolution for the whole terrestrial land surface between 56°S and 60°N, and results are validated against high‐resolution government flood hazard data sets from the UK and Canada. The global model is shown to capture between two thirds and three quarters of the area determined to be at risk in the benchmark data without generating excessive false positive predictions. When aggregated to ∼1 km, mean absolute error in flooded fraction falls to ∼5%. The full complexity global model contains an automatically parameterized subgrid channel network, and comparison to both a simplified 2‐D only variant and an independently developed pan‐European model shows the explicit inclusion of channels to be a critical contributor to improved model performance. While careful processing of existing global terrain data sets enables reasonable model performance in urban areas, adoption of forthcoming next‐generation global terrain data sets will offer the best prospect for a step‐change improvement in model performance.

show abstract

Evaluation of High-Resolution Precipitation Estimates from Satellites during July 2012 Beijing Flood Event Using Dense Rain Gauge Observations

Cited by 47 publications

References 31 publications

Investigating the Potential of Using Satellite‐Based Precipitation Radars as Reference for Evaluating Multisatellite Merged Products

Investigating the Potential of Using Satellite‐Based Precipitation Radars as Reference for Evaluating Multisatellite Merged Products

Comprehensive Evaluation of Two Successive V3 and V4 IMERG Final Run Precipitation Products over Mainland China

A high-resolution global flood hazard model

Contact Info

Product

Resources

About