Abstract
AbstractA method has been developed for estimating surface precipitation rates using reflectivities measured with a radar scanning at several different angles of elevation. In each pixel of a radar image, an idealized profile of the reflectivity factor is constructed. Each profile is defined in terms of three unknowns: the reflectivity in the rain beneath the melting layer, and the slope of the profile of the reflectivity in each of two layers above the melting layer. In these layers above the melting layer, it is necessary to invoke assumptions of horizontal homogeneity in the shape of the profile. Simple parametrizations are used for low‐level orographic growth and for the bright band, and the profile is diagnosed with the help of non‐radar information. the difference between the idealized and observed profiles is expressed as a single value for an entire radar‐image; the difference is minimized by iteration and this avoids using complicated methods of inversion. A simulation experiment was carried out for 15 different cases; in most of them, the incorporation of data from several elevation angles reduced bias errors at long range.The new method was then used on real radar‐data. In typical frontal rainfall, it showed no consistent improvement. Because it avoids using corrections which can otherwise become detrimental at long ranges, however, the new method gave improved results when reflectivity profiles were atypical.Unresolved spatial variability of the reflectivity profile continues to be a problem.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call