The ground trashing of ten day tamsat and CPC NOAA rainfall estimates over the western region of Uganda

Abstract

Rainfall products from operational algorithms like the TAMSAT and the NOAA CPC African Rainfall Estimate are being used for many diverse meteorological, climatological, hydrological, agricultural, and other applications. It is therefore important to have an idea of their accuracy and expected error characteristics. This is done by validating the satellite precipitation estimates against "ground truth" from rain gauge observations. One of the biggest challenges in generating real time climate information for decision support is monitoring rainfall performance over the western region of Uganda, which is characterized by a very sparse rainfall observation station network that has continued to degenerate over the past years. This has resulted into marked inconsistencies and gaps in gauge data. This region has a marked spatial and temporal rainfall variability which is attributed to its raised terrain, which is characterizing the western highland region together with the western albertine rift region. The highly variable rainfall in the tropics directly and indirectly affects different socio-economic sectors including agricultural production, water resources management, transport, construction, crop pest monitoring, disaster management, oil exploration among others. It is therefore essential to measure areal rainfall on real time basis in order to see the influence of the variable rainfall on such sectors. Remote sensing using satellites has been found very efficient for this as compared to the in-situ methods of observation. The comparison of gauge rainfall totals with the satellite estimates, whether on monthly or ten day temporal scales, remains an essential tool in the validation of satellite products, which may be used to augment the data gaps or supplement the available scanty data. This study is therefore to compare satellite derived rainfall estimates from two different operational algorithms, the NOAA CPC RFE and TAMSAT RFE with gauge data from five stations in western Uganda, with the main objective of presenting a ground truth and to show the expected errors in the estimates of the two algorithms. The study shows that both algorithms are in agreement with the rainfalt climatology patterns for all the five stations in the area of study. The TAMSA T RFE and the CPC RFE over Arua station showed the highest bias of (15.62 mm) and (7.96 mm) respectively. For both algorithms, Kasese station showed the least bias (-0.71 mm) for the TAMSAT RFE and (0.87mm) for the CPC RFE. The comparison further shows that the NOAA CPC RFE reflects a better agreement with the gauge data as compared to the TAMSAT RFE over the western region of Uganda. The NOAA Cf'C RFEs exhibited smaller biases as compared to the TAMSAT RFEs. Generally both RFEs exhibit a tendency to under estimate precipitation for most dekads within the study period for the five stations.