Differential propagation phase shift and rainfall rate estimation

This paper presents two methods for computing differential propagation phase shift (ϕDP) using time series data form a coherent radar with alternate switching between two linear but orthogonal polarizations. An analysis of the statistical error in ϕDP estimate shows that ϕDP can be estimated with less than 0.5° standard error using time and range averaging. To evaluate the usefulness of ϕDP for estimating rainfall rate (R) vis-a-vis the ZDR method, a discussion on the sensitivity of R to standard errors in ϕDP as well as ZH and ZDR is presented. It is shown that the relation, differential propagation phase constant Δϕ versus R, is relatively insensitive to drop size distribution (DSD) variations and thus can yield more accurate R estimate, even when a direct relationship between Δϕ and R is assumed. However, standard error in Δϕ estimate causes large inaccuracies of R at low values (R < 50 mmh−1), thus limiting its usefulness to higher rainfall rates. It is also shown that Δϕ can be used as a third remote measurable in conjunction with ZH and ZDR to determine a three parameter gamma DSD. Another use of Δϕ is likely to be in hydrometeor type identification, especially hail in severe storms.