SENSITIVITY OF THE ESTIMATED MONTHLY CONVECTIVE RAIN FRACTION TO THE CHOICE OF Z-R RELATION

Citation
M. Steiner et Ra. Houze, SENSITIVITY OF THE ESTIMATED MONTHLY CONVECTIVE RAIN FRACTION TO THE CHOICE OF Z-R RELATION, Journal of applied meteorology, 36(5), 1997, pp. 452-462
Citations number
81
Categorie Soggetti
Metereology & Atmospheric Sciences
ISSN journal
08948763
Volume
36
Issue
5
Year of publication
1997
Pages
452 - 462
Database
ISI
SICI code
0894-8763(1997)36:5<452:SOTEMC>2.0.ZU;2-A
Abstract
This study investigates the sensitivity of the estimated monthly conve ctive rain fraction-that is, the percentage of the areal rain accumula tion contributed by precipitation identified as convective-to variatio ns of the Z-R parameters used in radar-based rainfall estimation. Accu rate knowledge or the fractions of precipitation that are convective a nd stratiform is important for climatological studies estimating the h ealing of the atmosphere. Extensive datasets from two climatologically different precipitation regimes, Darwin, Australia, and Melbourne, Fl orida. are used. The potential uncertainty of using (if an arbitrary c hoice of the power factor b and (ii) either single or multiple Z-R rel ations (stratified by precipitation type) for converting radar reflect ivity to rain rate is investigated quantitatively. The analyses reveal that estimates of the monthly convective rain fraction are sensitive to the choice of Z-R parameters. A maximum sensitivity is found for pr ecipitation regimes with an approximately equal mix of rainfall from c onvective and stratiform precipitation systems. For example, estimates of the convective rain fraction for monsoonal rainfall at Darwin may range from 30% to 80% solely depending on the choice of Z-R parameters , even though all of these Z-R relations are tuned to produce the same total rainfall. In contrast, for the highly convective, sea-breeze-tr iggered, multicellular storms around Melbourne. the estimates of the c onvective rain fraction may range from 80% to 100% Different approache s to how the appropriate parameters of the Z-R relation(s) may be obta ined are discussed. Varying the Z-R parameters to maximize the correla tion of the radar-estimated monthly rainfall at the gauge sites and th e rain gauge accumulations does not reveal enough sensitivity to make any choice significantly better than a single Z-R relation for both co nvective and stratiform rain. Multiple Z-R relations may be justified, but apparently nor on the basis of a convective-stratiform separation .