A high spatiotemporal gauge‐satellite merged precipitation analysis over China. Issue 6 (24th March 2014)

• Record Type: Journal Article
• Authors: Shen, Yan; Zhao, Ping; Pan, Yang; Yu, Jingjing
• Is Part Of: Journal of geophysical research. Volume 119:Issue 6(2014:Jun.)
• Page Start: 3063
• Page End: 3075
• Publication Date: 2014-03-24
• Subjects: Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1002/2013JD020686
• Languages: English
• ISSNs: 2169-897X

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <p>Using hourly rain gauge data at more than 30, 000 automatic weather stations in China, in conjunction with the Climate Precipitation Center Morphing (CMORPH) precipitation product for the 2008–2010 warm seasons (from May through September), we assess the capability of the probability density function–optimal interpolation (PDF‐OI) methods in generating the daily, 0.25° × 0.25° and hourly, 0.1° × 0.1° merged precipitation products between gauge observations and the CMORPH product. We find that error correlation, error variances of gauge and satellite data, and matching strategy in the PDF‐OI method are dependent on the spatial and temporal resolutions of the used data. Efforts to improve the parameters and matching strategy for the hourly and 0.1° × 0.1° product have been conducted. These improvements are not only suitable to a high‐frequency depiction of no‐rain events, but accurately describe the error structures of hourly gauge and satellite fields. The successive merged precipitation algorithm or product is called the original PDF‐OI (Orig_PDF‐OI) and the improved PDF‐OI, respectively. The cross‐validation results show that the improved method reduces systematic bias and random errors effectively compared with both the CMORPH precipitation and the Orig_PDF‐OI. The improved merged precipitation product over China at hourly, 0.1° resolution is generated from 2008 to 2010. Compared with the Orig_PDF‐OI, the improved product reduces the underestimation greatly and has smaller bias and root‐mean‐square error, and higher spatial correlation. The improved product can better capture some varying features of hourly precipitation in heavy weather events.</p> </abstract>