Análise da atenuação do sinal SAR na banda C por precipitação: estudo de caso na ilha do Mosqueiro/Belém – PA

Abstract

Passive and active sensors play distinct roles in the scenario using sensors to perform mapping. Passive sensors work in the visible and infrared range, while active sensors operate in the microwave range of the electromagnetic spectrum. However, passive sensors face limitations, such as dependence on favorable weather conditions to obtain data and the inability to operate at night, in addition to interference from clouds in regions of high cloudiness, such as the Amazon. The objective is to investigate how the attenuation caused by rain affects the Radar signal (radio detection and ranging) and its behavior on targets. This contributes to better understanding the performance of these systems in adverse weather conditions. The work methodology sought to analyze the interference caused by rain in the SAR sensor on Mosqueiro Island, three classes were analyzed, which were called forest, exposed soil and pasture and for this purpose Google Earth Engine was used to process the images and obtain the values of backscatter in the VV and VH polarization of each sample referring to the three classes. To separate the classes, the stratified random sampling design criterion was used, a data collection method in which the population is divided into subgroups called strata. Furthermore, the Jamovi software was used to analyze the data and through it it was possible to carry out a statistical analysis between the class samples. The method used to carry out the analysis was Kruskal – Wallis, which is a non-parametric method used to analyze whether there are statistically significant differences between groups. Multiple comparison tests can be used to identify which groups differ from each other. This statistic is then compared to a reference distribution to determine if there are statistically significant differences between the groups. The results achieved showed that the SAR sensor undergoes changes resulting from rain and its behavior on the surface. Added to this, all classes analyzed showed significant results in the value of the statistics applied to analyze the data. It was noted that there was a more significant difference in VV polarization than in VH due to attenuation from rain and humidity. Therefore, it is clear that rainfall in active sensors must be taken into account when mapping in areas with high rainfall.