Recently, it has been shown that country-wide rainfall maps can be obtained from microwave link data for the Netherlands (35,000 square kilometers). Rainfall maps solely based on received signal levels compare quite well with those based on gauge-adjusted radar data for a 12-day validation dataset. Hence, rainfall estimation using microwave links holds a promise, although more research is needed to obtain accurate rainfall maps all year round, i.e., to develop a real-time application.
For countries operating weather radars a potential benefit of link rainfall data lies in merging with radar rainfall data. Often radar rainfall images are adjusted employing rain gauge data to improve the quality of rainfall maps. However, the number of automatic rain gauges suitable for adjustment is generally low, thus limiting the improvement of real-time radar rainfall images of, e.g., 1-hour rainfall depths. In general, the number of microwave links in a country will be much larger than the number of (automatic) rain gauges. E.g. the Royal Netherlands Meteorological Institute operates 32 automatic rain gauges in the Netherlands (1 gauge per 1000 square kilometers), whereas data from 1500 link paths out of 8000 are available. Microwave links are not expected to be as accurate as rain gauges, but their large number may partially compensate for this.
The potential of microwave link data to improve real-time radar rainfall products is investigated. A dataset from a commercial microwave link network over the Netherlands is analyzed, containing data from an unprecedented number of links (~ 1500) covering the land surface of the Netherlands (35500 square kilometers). This dataset spans from January 2011 through October 2012. Hourly rainfall data are derived from the microwave link data and employed to adjust hourly country-wide radar rainfall maps, which are accumulated to daily rainfall maps. The daily rainfall maps are validated against a high-quality, climatological, gauge-adjusted radar rainfall dataset and will also be compared to the existing real-time radar rainfall product which has been adjusted using data from 32 automatic rain gauges.
Adjustment of radar data utilizing microwave link data could substantially improve the quality of real-time radar rainfall products and thus increase their applicability for hydrological purposes, validation of precipitation forecasts of numerical weather prediction models, as well as validation of satellite-based rainfall retrievals.