Performance Analysis of Rain Rate Conversion Models of 60-Minute to 1-Minute Integration Time for Microwave Applications Over Subtropical Coastal Regions

The expansion and utilization of higher frequency bands are now the focus of radio-communication systems engineers globally, including in South Africa. Rain-induced attenuation has been identified as the critical factor degrading signals, especially at frequencies above 10 GHz, with greater impact in areas with high-intensity precipitation year-round. 1-minute rain rate data are needed as an input to predict rain-induced attenuation and propose a suitable mitigation measure, as recommended by the International Telecommunication Union. However, 1-minute rain rate data are not readily available worldwide due to the cost of installing equipment at different locations, since rainfall is location-dependent. However, rain data with higher integration times are readily available, especially via satellite measurements. The paper focuses on converting the available higher integration time (60-minute) to a 1-minute rain rate using appropriate conversion techniques. This paper reviews rainfall distribution and evaluates the performance of 60- to 1-minute rain-rate integration models over coastal areas of South Africa. We chose five models based on prior performance in the tropical region and evaluated their performance using root mean square error (RMSE) and relative error at the chosen locations. The results show that the Burgueno model performs best over Durban, while three different models (Moup and Martins; Segal; and Emiliani’s models) yield lower RMSE values in different cases over Cape Town. The Burgueno model yields lower values than Durban for a crucial percentage of the time, while the LG model achieved only a 0.3% exceedance.