EVALUATION OF ERA-INTERIM AND ERA5 REANALYSIS ON THE DIURNAL CYCLE OF RAINFALL IN THE WESTERN PART OF JAVA

• Main Author: Restu Utari, Mega
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/70402
• Institution: Institut Teknologi Bandung

The diurnal cycle is one of the important factors in the occurrence of regional rain in Indonesia. Diurnal cycle patterns can vary in character in various temporal scales. Reanalysis is an alternative solution that can be used to study diurnal cycles in the long term due to the availability of data. Evaluation of reanalyzed data cannot be ignored because of its uncertainty. So the purpose of this study was to evaluate the ERA5 and ERA-Interim reanalysis data on the diurnal cycle, especially in Western part of Java. The evaluation was carried out by comparing the reanalysis data with surface observation data and satellite data with the best performance (between GSMaP and IMERG). Evaluation of reanalyzed data on surface observation data and satellite data was carried out at four observation stations (Pacet, Pengalengan, Observatorium, and Cibinong), then the correlation and RMSE values were calculated, and compared the results of the spatiotemporal plot with selected satellite data based on the composite value per 6 hours. Furthermore, the reanalyzed data and selected satellites were evaluated for the phenomenon of diurnal rainfall propagation in the northern part of Western part of Java based on monthly composites. The results show that in general the performance of ERA5 is better than ERA-Interim, this can be seen from the average value of diurnal rainfall ERA5 is closer to surface observations compared to ERA-Interim, where the value of ERA-Interim is overestimated in all stations. In terms of the composite seasonal average, ERA5 can capture the spatial pattern of diurnal rainfall which is more similar to the IMERG satellite. The statistical test results show that ERA5 has a higher average spatial correlation value of 0.88 and a lower average RMSE of 1.71 than the Interim ERA, respectively 0.69 and 2.42. ERA5 can capture diurnal rainfall propagation and early morning precipitation in northern Western part of Java with smaller values, while ERA-Interim does not.