Evaluation of precipitation from CMORPH, GPCP-2, TRMM 3B43, GPCC, and ITPCAS with ground-based measurements in the Qinling-Daba Mountains, China.

Gefei Wang,Liwen Liang,Peiyun Zhang,Shiqiang Zhang,Jeffrey Shaman

doi:10.1371/journal.pone.0185147

Abstract

The correspondence between five precipitation products, including CMORPH, GPCP-2, TRMM 3B43, GPCC, and ITPCAS, and ground-based measurements of precipitation were evaluated on annual, seasonal, and monthly scales during 2000–2014 in the Qinling-Daba Mountains over China, which is a significant area with vital value of climate and hydrology. Performances of the precipitation products in the relatively arid/humid years were also analyzed. In general, ITPCAS data displayed the highest accuracy, GPCP-2 and CMORPH data showed relatively poor performance, and GPCC and TRMM 3B43 data were average at different temporal scales among the five precipitation products. The Pearson correlation coefficient of each station had minor fluctuations for the five precipitation products. A larger deviation was found at Wudu station, most likely due to the undulating terrain. The performances of the precipitation products from highest to least accuracy are as follows: ITPCAS > TRMM 3B43 > GPCC > GPCP-2 > CMORPH. Except for CMORPH (-20.76%), the percentage precipitation differences (PPDs) of the other four precipitation products fluctuated in the range of 10% during the relatively arid (2001) and humid (2011) years. In addition, all precipitation products and ground gauge observed precipitation did not show an obvious gradient with altitude, which is different from that in other mountainous areas and is perhaps due to complex terrain, lack of observation in high altitudes, and precipitation undercatch. In consideration of the significance of Qinling-Daba Mountains as the geographic and ecological dividing lines, the present study may provide a new perspective for hydrological, climatic, and ecological researches and practices in local and other mountainous areas.

Highlights

Precipitation is widely recognized as a fundamental component of the global water cycle and plays a critical role in the survival of human beings as our primary source of fresh water
It is clear that TRMM 3B43, GPCC, and ITPCAS had high values of r above 0.8, which indicate that these three products, especially ITPCAS with an r value of 0.90, had high agreement with rain gauge observations on an annual scale
ITPCAS and GPCC performed better than GPCP-2with precipitation differences (PPDs) of 3.69% and 4.78%, respectively, and TRMM 3B43 and CMORPH were averaged between, which showed that the PPDs of all precipitation products were within 10%

Summary

Methods

Aggregated to monthly, seasonal and annual precipitationAggregated to seasonal and annual precipitation Aggregated to seasonal and annual precipitation Corrected and aggregated to seasonal and annual precipitation Aggregated to seasonal and annual precipitationCMORPH, a satellite-only product, is produced through a morphing technique that uses precipitation estimates from passive microwave observations and propagates these features using motion vectors from geostationary satellite IR (infrared) imagery at half-hour intervals [22]. For the global precipitation products with the finest resolution in this paper, CMORPH with a 0.07277 ̊ latitude/longitude resolution (8 km × 8 km at the equator) was obtained for analysis every 30 min between 60 ̊S and 60 ̊N (http://www.cpc.ncep.noaa.gov/products/janowiak/ cmorph_description.html). GPCP-2 has the coarsest resolution compared to the other four precipitation products analyzed in this study. It incorporates precipitation estimates derived from low-orbit microwave data, geosynchronous infrared data, and surface rain gauge observations. GPCP Version is a combination of precipitation estimates from MW (microwave) and IR (infrared) sensors in polar orbit satellites and geostationary satellites, as well as surface observations. For proper evaluation and comparison of precipitation data from different sources, a uniform spatial and temporal resolution was first acquired to place all precipitation products under the same conditions [14, 39].

Results

Discussion

Conclusion