Thesis Colloquium Tewodros Yate

Qualities of daily rainfall maps from X-band radar and microwave links rainfall retrievals were evaluated. The quality of 15-min link rainfall retrievals was also assessed. Furthermore, radar rainfall spatial correlation is also computed using season-long daily rainfall data.

Organised by Hydrology and Quantitative Water Management

Thu 12 March 2015 16:00 to 16:30

Venue Lumen, gebouwnummer 100
Room Lumen 2

Evaluating IDRA radar and microwave link rainfall retrievals using gauge-adjusted climatological radar rainfall data

All rainfall measuring techniques bring disadvantages with them and are not always adequately available in space and time. Hence, improved techniques and new alternatives to support rainfall monitoring are proposed. For instance, the high resolution X-band radar at Cabauw and rainfall estimation using microwave links from cellular communication networks. In its standard configuration, the X-band radar provides 30 m spatial resolution at 1 minute temporal resolution. Microwave link networks cover significant areas and measure closer to the earth’s surface compared to ground-based weather radars. This study evaluated the quality of daily rainfall maps from X-band radar and microwave link rainfall retrievals for a study area located in Cabauw, near Utrecht. It consists of a circular region of approximately 15-km radius, the coverage of the IDRA radar. The quality of 15-min link rainfall retrievals was also assessed. Moreover, the study explored rainfall spatial correlation at different accumulation intervals using season-long daily rainfalls from a radar rainfall dataset. Gauge-adjusted climatological radar rainfall and rain gauge rainfall data were employed to accomplish the objectives of this research. The quality of daily rainfall maps from the IDRA radar compared to the reference gauge-adjusted climatological radar was found to be unsatisfactory. Out of 14 days considered, only 5 days compared fairly well. The quality of 24-h rainfall maps from microwave links was found to be generally good, given the assessment was made for a local rural area with a limited link density. Out of 8 days assessed, maps of 6 days showed a reasonable agreement. Time series analysis showed that the dynamics of 15-min link-based rainfall depths generally corresponded well with that of gauge-adjusted radar and rain gauge rainfall. The radar rainfall depths showed a spatial correlation pattern for summer 2011 for all accumulation intervals except 24 h, whereas no clear spatial correlation pattern could be found for winter 2012.