Cloud Properties Observed by Ground-Based Millimeter Wavelength Radar in the Korean Peninsula

The cloud observations from ground-based millimeter wavelength radar provide vertical distribution of cloud layers with good accuracy and high vertical resolution and can be utilized to produce cloud type climatologies. We examine statistical cloud properties according to the cloud types using Ka-band cloud radar (KCR) that are operated by National Institute of Meteorological Research (NIMR) of Korea Meteorological Administration (KMA) since 2013. For stable and accurate statics of cloud information, strong echoes from insects in the boundary layers were removed and system biases in measurements from KCR were calibrated. We, thus, developed an algorithm for identifying and eliminating of insect-contaminated echoes using Linear Depolarization Ratio (LDR). The probability density functions of three feature parameters (LDR, texture of LDR, and degree of spatial spareness for LDR) were utilized to detect and remove the insects-contaminated echoes. The calibration of system biases in reflectivity from KCR was then conducted by using observations from collocated external instrument. The system biases in KCR reflectivity were corrected by direct comparison with reflectivity from vertically pointing X-band radar (VertiX) that are pre-calibrated by using drop size distributions from two-dimensional video disdrometer (2DVD) for rain events. For the analysis of cloud climatology, cloud properties (i.e., cloud fraction, top and base heights of the cloud, and cloud thickness) were examined according to the cloud types that were objectively classified with their altitude, presence of precipitation, and vertical extents in vertical profiles of reflectivity. First, clouds were categorized into high, middle and low clouds based on height of cloud base. The low clouds were then divided into “rain” and “non-rain” based on presence of precipitation. “rain” and “non-rain” clouds were classified into deep and shallow clouds by their vertical extents. Preliminary analysis shows that the low cloud was dominant in comparison with the middle and high clouds for the period of September and October in 2013. The fraction of deep clouds without precipitation was greater than the one of shallow clouds without precipitation.

Acknowledgements This research was supported by a grant (14AWMP-B079364-01) from Water Management Research Program funded by Ministry of Land, Infrastructure and Transport of Korean government.