<br />388
<br />
<br />PJ. BREMAUD AND Y.B. POIr-rrIN
<br />
<br />FORECASTING HEAVY RAINFALL FROM RAIN CELL MOTION
<br />
<br />389
<br />
<br />Andrieu, H.. Creutin. J.D., Delrieu, G.. leoussof, J. and Pointin, Y. 1989. Radar data
<br />processing for hydrology in the Cevennes region. IAHS II1rd Scientific Assembly, 10-19
<br />May 1989, Baltimore, IAHS Publication No. 186, pp. lOS-I 15.
<br />
<br />Austin, G.L. and Bellon, A., 1914. The use of digital weather radar records for short.term
<br />precipitation forcasting. Q. J. R. Meteorol. Soc., 100: 658-664.
<br />Barclay, P.A. and Wilk, K.E. 1970. Severe thunderstorm radar echo motion and related
<br />weather events hazardous to aviation operations. ESSA Tech. Memo. 46. National Severe
<br />Storms Laboratory, Nonnan, OK., 63 pp.
<br />Bellon, A. and Austin. G.L. 1978. The evaluation of two years of rcal-time operation of a
<br />short-teon precipitation forecasting procedure (SHARP). J. Appl. Meteorol., 17: 1778-1787.
<br />Bjerkaas, C.J. and Forsyth, D.E. 1980. An automated real-time storm analysis and storm
<br />tracking program (WEATRK). Contract AFGL-TR-80-0316, Air Force Geophysics
<br />Laboratory, Hanscom AFR, MA, 52 pp.
<br />Blackmer, Jr. R.H., Duda, R,O. and Reboh, R., 1973. Application of pattern recognition
<br />te<::hniques to digitized weather radar data. Final Report, Contract 1-36072, Stanford
<br />Research Institute, Menlo Park, CA, 89 pp.
<br />Bremaud, P.J., 1991. Suivi et prevision automatiques du deplacement des nuages precipitants
<br />par radar meteorologique. Ph.D. Thesis, University Blaise Pascal, Clermont-Ferrand, 202
<br />pp,
<br />Brunkow, O.A., 1980. A digital radar-based rainfall monitoring and forecasting tool. 19th
<br />Conference on Radar Meteorology, 15-18 April 1980, Miami, American Meteorological
<br />Society, Boston, MA, pp. 62-65.
<br />Collier, e.G., 1978. Objective forecasting using radar data: a review. Meteorological Office
<br />Radar Research Laboratory, Malvern, UK, Res. Rep. 9, 29 pp.
<br />Collier, C.G., 1989. Short-period forecasting using radar data. In e.G. Collier (Editor),
<br />Applications of Weather Radar Systems. A Guide to Uses of Radar Data in Meteorology
<br />and Hydrology. Ellis Horwood, Chichester, UK, pp. 136-182.
<br />Crane, R.K., 1979. Automatic cell detection and tracking. IEEE Trans. Geosci. Electron.,
<br />GE-11-(4): 250-262,
<br />Einfalt, T., 1991. Summer 1990 in Seine-Saint~Denis: forecast versus measured areal rainfall
<br />values. Atmos. Res., 27: 1.51-158.
<br />Einfalt, T., Denoeux, T. and Jacquet, G., 1990. A radar rainfall forecasting method designed
<br />for hydrological purposes. J. Hydrol., 114: 229-244.
<br />Elvander, R.C., 1976. An evaluation Qf the relative performance of three weather radar e<::ho
<br />forecasting techniques. 17th Conterence on Radar Meteorology, 26-29 October 1976,
<br />Seattle, American Meteorological Society, Boston, MA, pp. 526-532.
<br />Huff, F.A., Changnon, S.A. and Vogel, J.L., 1980, Convective rain monitoring and forecasting
<br />system for an urban area. 19th Conference on Radar Meteorology, 26-29 October 1976,
<br />Miami, American Meteorological Society, Boston, MA, pp. 56-61.
<br />Messaoud, M. and Pointin, Y.B., 1990. Small time and space measurements of the mean
<br />rainfall rate made by a gage network and by a dual polarization radar. J. Appl. Meteorol.,
<br />29: 830--841.
<br />Wilk, K.E. and Gray, K.C., 1970. Processing and analysis techniques used with the NSSL
<br />weather radar system. 14th Conference on Radar Meteorology, 17-20 November 1970,
<br />Tucson, AZ, American Meteorological Society, Boston, MA, pp. 369-374.
<br />Wilson, J.W. and Kessler, III, E., 1963. Use of radar summary maps for weather analysis and
<br />forecasting. J. Appl. Meteorol., 2: 1-11.
<br />Zittel, W.O., 1976. Computer applications and techniques for storm tracking and warning.
<br />17th Conference on Radar Meteorology, 26-29 October 1976, Seattle, American Meteoro-
<br />logical Society, Boston, MA, pp. 514-521.
<br />
<br />a new forecasling method (Bremaud, 1991) which focuses on heavy rainfall,
<br />as it detects and exlrapolales the molion of heavy rainfall cells which are
<br />closely associated with conveclive cells, Indeed, both the precipitation fore-
<br />casting method and the radar data must be adapled 10 their use, parlicularly
<br />in terms of the desired forecast period, because the meteorological structures
<br />associated with rain can have very different spatial and temporal characteris-
<br />tic scales: from I km and 30min for the convective cells to 1000 km and 48h
<br />for the extratropical cyclones, each of these different scale structures having
<br />their OWn motion, Therefore, to oblain the best rainfall forecast, Ihe method
<br />must detect and extrapolate the motion of the precipitating meteorological
<br />structure whose duration is about the forecasting time interval and whose
<br />motion can be detected on the successive radar pictures. In the case of a I h
<br />forecasting time interval, the most suitable melhod is that which uses small
<br />spatial scale radar data and extrapolates the molion of rain cells associated
<br />with convective cells.
<br />We have compared our method (named 'P ARAPLUIE') with three other
<br />methods, one of which (SCOUT 11.0) is at present used for sewer
<br />management. The results according to the 'concordance' or hydrological
<br />criteria show that our method is reliable and performs well whatever the
<br />studied event or the forecasting time interval used, up to 15 min. Thus, the
<br />PARAPLUIE method appears to be well suited for forecasling time intervals
<br />of aboul15 min to I h, whereas the cross-correlation methods perform belter
<br />in the case of forecasting time intervals of 2 h or more by using composite
<br />radar data, For 1-2 hours lead time, the most suited method may be that
<br />which tracks the slructures enclosing the convective cells, i,e, the method
<br />extrapolating T echoes defined at a low reflectivity threshold (aboul 20 dB),
<br />
<br />ACKNOWLEDGEMENTS
<br />
<br />The authors are indebted to Dr. T, Einfalt for discussions and for making
<br />the radar data of the Paris 1989 event available, The Cevennes 1988 and the
<br />other radar data have been obtained by the members of the radar group of the
<br />LaMP/OPGC and by all campaign participants, The mathematical tools,
<br />such as the pattern recognition techniques which are used in the
<br />PARAPLUIE method, have been described to us by R, Pejoux and 0,
<br />Chambreuil. The edition of this manuscript owes much to Dr, H, Lassen and
<br />0, Guillot and 1. Squarise, We express our thanks to all these persons,
<br />
<br />REFERENCES
<br />
<br />
|