Cloud detection and attenuation modelling in tropical region

• Main Author: Yuan, Feng
• Other Authors: Lee Yee Hui
• Format: Theses and Dissertations
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Satellite telecommunication
• Online Access: http://hdl.handle.net/10356/74836
• Institution: Nanyang Technological University

Cloud plays a significant role in both the ground-to-space radio communication system and climate system of Earth. In a tropical country like Singapore, more than 80% of observed clouds are thick cumulus/cumulonimbus clouds that are heavy with high liquid water content, and consequently accounts for a significant amount of attenuation on satellite signals. In this thesis, various types of data are collected from different kinds of instruments including radiosonde, ceilometer, dual-polarization weather radar, Ka-band beacon receiver, weather station and whole sky imager. By processing and analyzing these data, we study the various aspects of cloud including cloud vertical structure, melting layer of rain cloud, and the impact of cloud on the satellite communication for tropical region. To detect cloud vertical structure, four year radiosonde data from 9 tropical stations are processed and analyzed. A new method using water vapor pressure (WVP) is then proposed for cloud detection in the tropical region. The proposed model detects the presence of clouds based on the following criterion: the measured water vapor pressure is larger than the derived critical water vapor pressure at the same level. The estimated cloud vertical structure using the proposed method is compared with the Salonen and Uppala (SU) model, the ceilometer data and two kinds of meteorological observation data, namely, SYNOP (surface SYNOPtic observations) and METAR (MEteorological Terminal Aviation Routine weather report) . Our proposed model shows a higher accuracy of prediction of the cloud vertical structure as compared to the existing SU model. After studying the cloud vertical structure, the melting layers for the stratiform and convective rain clouds are investigated by using data from the dual-polarization radar. In the tropical region, convective rain dominates the rain events, however, knowledge of the melting layer from convective rain is very limited. Therefore, we processed and analyzed the dual-polarization radar data to identify both stratiform and convective rain melting layer in the tropical region of Singapore. By studying and differentiating these two types of rain events, it is found that amongst the three radar measurements of reflectivity (Z), differential reflectivity (ZDR), and cross-correlation coefficient (ρHV ), the best indicator for the melting layer of convective rain events is the cross-correlation coefficient. After comparing the convective rain melting layer top heights withe the corresponding 0 °C isotherm heights, it is found that for convective rain events, the threshold to detect melting layer should be modified to ρHV = 0.95 for tropical region. In order to detect clouds, several cloud detection methods are applied in order to achieve accurate cloud attenuation estimation. Two year statistical beacon data are collected and processed. Experimental results from the beacon receiver is also analyzed for the calculation of cloud attenuation. These results are then compared to the existing ITU-R model for same frequency and elevation angle. The complementary cumulative distribution function (CCDF) of the 2-year cloud attenuation derived from the beacon receiver shows that the Ka-band cloud attenuation can be up to 4 dB at the time exceedance of 0.01%. The ITU-R model is found to underestimate the cloud caused attenuation in the tropical region. After analyzing the cloud attenuation along the satellite slant path link, a new model is proposed to improve the ITU-R cloud attenuation model. The improved model incorporates precipitable water vapor (PWV) value to estimate integrated liquid water content (ILWC), and then determine the cloud attenuation for the tropical region. From the analysis of 2 year radiosonde data from 8 cites in the tropical region, the results show that the ILWC along the path can be approximated by a power function relationship with the PWV. The estimated cloud attenuation using the improved model is then compared with the values calculated using the existing ITU-R model and the cloud attenuation derived from a Ka-band beacon data. Results calculated from the proposed model has shown good agreements with the ITU-R model at high percentage of time exceedance which also matches well with the cloud attenuation suffered by the beacon signal at low percentage of time exceedance. It is noted that with the introduction of the use of PWV data from GPS (Global Positioning System), higher temporal resolution of cloud attenuation can be achieved comparing to the existing ITU-R cloud attenuation model. In this thesis, we have proposed a model to detect cloud vertical structure, studied and detect the melting layer for rain cloud, done statistical analysis of cloud attenuation on Ka-band satellite link and also proposed a method for cloud attenuation calculation in tropical region. The algorithms and techniques developed in this thesis will assist in remote sensing applications and satellite communication applications.