Received: 18 Mar 2015 – Accepted for review: 18 Mar 2015 – Discussion started: 22 Apr 2015
Abstract. There are many potential sources of bias in the radar rainfall estimation process. This study classified the biases from the rainfall estimation process into the reflectivity measurement bias and QPE model bias and also conducted the bias correction methods to improve the accuracy of the Radar-AWS Rainrate (RAR) calculation system operated by the Korea Meteorological Administration (KMA). For the Z bias correction, this study utilized the bias correction algorithm for the reflectivity. The concept of this algorithm is that the reflectivity of target single-pol radars is corrected based on the reference dual-pol radar corrected in the hardware and software bias. This study, and then, dealt with two post-process methods, the Mean Field Bias Correction (MFBC) method and the Local Gauge Correction method (LGC), to correct rainfall-bias. The Z bias and rainfall-bias correction methods were applied to the RAR system. The accuracy of the RAR system improved after correcting Z bias. For rainfall types, although the accuracy of Changma front and local torrential cases was slightly improved without the Z bias correction, especially, the accuracy of typhoon cases got worse than existing results. As a result of the rainfall-bias correction, the accuracy of the RAR system performed Z bias_LGC was especially superior to the MFBC method because the different rainfall biases were applied to each grid rainfall amount in the LGC method. For rainfall types, Results of the Z bias_LGC showed that rainfall estimates for all types was more accurate than only the Z bias and, especially, outcomes in typhoon cases was vastly superior to the others.
How to cite. Lee, J.-K., Kim, J.-H., and Suk, M.-K.: Application of bias correction methods to improve the accuracy of quantitative radar rainfall in Korea, Atmos. Meas. Tech. Discuss., 8, 4011–4047, https://doi.org/10.5194/amtd-8-4011-2015, 2015.
This paper focuses on correcting the reflectivity measurement bias and the QPE model bias. For the reflectivity bias correction, the bias correction algorithm which corrected target radars in the hardware and software bias based on the standard radar was utilized. The Mean Field Bias Correction and the Local Gauge Correction methods were also applied to correct the QPE model bias. Proposed bias correction methods are able to contribute to the real-time QPE model in work-site operation.
This paper focuses on correcting the reflectivity measurement bias and the QPE model bias. For...