Radar vertical profiles and melting layer studies from an S-band Doppler polarisation-diversity radar campaign in Singapore

As a part of the TRMM and the EuroTRMM projects to provide ground validation, the Radio Communications Research Unit at the Rutherford Appleton Laboratory has designed and built an S-band Doppler polarisation-diversity radar, operational since 1998 in Singapore. Four parameters are measured by the system; radar reflectivity (Z), linear depolarisation ratio (LDR), Doppler velocity and spectrum width. In addition, data from radiosonde measurements are also available from Singapore.

Information regarding the vertical structure, the melting layer height and thickness in tropical regions has been sparse due to lack of experimental data. The melting layer is the transition region between snow and rain. It usually starts at the 0oC isotherm level and finishes a few degrees above 0oC, when all the particles have melted and become raindrops. The information about the melting layer height and thickness, and the vertical profile structure is vitally important to remote sensing because it can identify the raincell height and the impact of the melting particles. These factors are important to the TRMM rainfall retrieval algorithms.

The observation of the melting layer using a variety of radar parameters gives rise to bright band signatures, which are caused by a high concentration of particles of high index of refraction arising from the melting snow, with the additional factor of non-spherical shapes. Based on the vertical profiles recorded by the radar during stratiform events over a three year period, we have derived the top and bottom of the melting layer, resulting in the thickness of the melting layer. Comparisons have been made between the radar profiles and the 0oC isotherm height measured by radiosonde.

To study the variation in radar reflectivity vertical profile, we have extracted several points of interest at different heights (close to the ground, middle of rainfall, near bottom of melting layer and near to the top of melting layer). As a result, we have studied the systematic variations and the relationship between ground rainfall, melting height and thickness, and in particular, the trend of variation, namely increasing or decreasing with height. The non-uniformity in radar vertical profiles may be caused by the changes in raindrop size distribution, and probably evaporation.