Assessment and Calibration of Gilan Radar Precipitation intensity using ground station data

Radar is a remote sensing instrument that sends electromagnetic waves with specific power to atmosphere and evaluates the amount of return power. It can then measure the difference between the send and retune powers and detect atmospheric phenomena as clouds. Using this tool, there is a wide, continuous and integrated monitoring of atmospheric phenomena. Like any remote sensing device that has, data of weather radars can also have errors. One of the most important measures to eliminate or minimize the radar data errors is calibration, and correction of radar index coefficients. The purpose of this paper is to extract an appropriate relationship for precipitation intensity related to radar reflectivity in Gilan. The Gilan ground based radar installed at the Kiashahr Port is a German-made GEMATRONIK MTEOR1600C type operating in the dual-polarization Doppler radar frequency band (c-band). In general, in order to calibrate the weather radars, “a” and “b” coefficients are required to modify in the Marshall Palmer initial formula for the target area. For this purpose, we tried to estimate the coefficients of this relationship (the relationship between precipitation intensity and radar reflection intensity) in a three years period (2012-2015) and to find new coefficients. In this study, the Doppler filter method (IIR Doppler Filter) was used to remove clutters. This filter was installed in the signal processor. In order to calibrate the Gilan radar, the rain gauge data of the Rasht airport synoptic station was compared with radar data. In this way, the precipitation statistics of the meteorological station were extracted using available meteorological and scdata software in the selected period and were separated based on two views of the season and precipitation severity.
Then the precipitation intensity was calculated based on radar data. Due to the large amount of raw data in Rainbow software, the data format was converted from binary to text. In the next step, the power regression is made between meteorological radar data ( ) and the automatic rain gauge (in mm / h), based on the existing default coefficients. Then, the new coefficient (a’ and b’) were determine by introducing the linear equation, a (a') and a new b (b'). In the third step, the precipitation intensity was re-calculated by applying new coefficients in radar measuremets. Now, there are 2 precipitation intensity values which are obtained by default and new coefficients. The intensity precipitation values were compared with observation of meteorological data using root mean square error in different seasons regardless of intensity. The same process was performed for the severity of observed precipitation and calculated precipitation by the radar regardless of seasons. The most important results of this study are relative improvement of radar estimation from precipitation intensity after correction of coefficients, which was 38% in March to May (spring) and 22% in December to February (winter). During the months of July to November (summer and autumn), there was almost no improvement. Also, based on precipitation intensity (regardless of seasons), average accuracy of precipitation was increased by 25% and severe precipitation by 47%. While in gentle precipitation, this method did not work and there was no improvement