A Modification of the Koppen Criteria for Determining Seasonal Distribution of Precipitation

A Modification of the Koppen Criteria for Determining Seasonal Distribution of Precipitation ANNA MARIE BOSOHEN University of California, Berkeley According to the Koppen climatic classification , tho differentiation of humid and dry climates is dopendent upon the three variables: procipitation regime, total annual precipitation, and moan annual temperature. In applying all three of thoso variables in a study of climatic variability in the western United Statos, it was found that tho Koppen criteria for precipitation regime were unsatisfactory. The dopondenco of the regime upon the amounts of precipitation of a single month in each of tho two seasons means that the oceurronco of little or no rain in one winter month and in ono summer month automatically classos tho station as having a uniform regime, irrespective of tho seasonal totnls. R. J. Russell, in a study oí tho dry climates of the United Statos published in tho University of California Publications in Geography, volume 5, 1D31, prisscntod a modification of these criteria, This is, however, too elaborate for uso in connection with climatic variability since, according to it, a station may bo classified as any ono of twenty-seven different types in successive years. Russell , in his method, considers only two of the three months of each season because he fools that the use of all three would give too much weight to erratic months. ?? examination of climatic records readily shows that the consideration of only two months instead of three will give greater weight to an "erratic" month. The record of Ricardo, Texas, will illustrate this point. Jan. 1,15 June 3 44 Feb. 0.83 Aug. 1.06* "L98TJO Dw. 1.1« June 3,34 ,Inn. 1.1 5 July 1,90 Feb. 0.S3 Aug. 1,06 3.14' 7.30 Since tho summer is the wetter season, one would take the menu of the two wettest summor months and of tho two driest winter months, according to Russell, as a basis for determining tho precipitation regime. Accordingly, the two "typical" or driest winter months would be January and February while Docomher, which lins only 0.01 incites more than January, is loft out of consideration. Similarly, ,TnIy with 0.06 inches loss than August is omittod in preference to June with 1.54 inches moro tlifln July. Certainly, this approach givos too much weight to an erratic month. In the Geographical Review for Jann firy, 1935, H. M. Kendall, in an article on climatic boundaries of tho eastern United States, suggests that summer concentration can bo said to occur when 70% of the precipitation falls during tho summer six months and likewise 70% falling in the winter six months will give a winter concentration. No indication is given that these values aro othor than arbitrarily chosen and their significance must ho questioned .since no consideration is given to the relativo effectiveness of winter and summer precipitation. The most accurate statement of the sensonnl distribution of precipitation Is certainly the one which takes into account the total precipitation of the seasons, In order to find a suitable, set of ratios to .uso with seasonal totals, tho records in* about ono thousand stations wore plotted. Tho ratios of single months wore used as abscissae' and the ratios of the seasonal totals or ordinates. Separate graphs wore made for stations with dry summer regimes and thoso with dry winter regimes, Tho ratios thus empirically determiner! were 1.5 for a dry summer regime, instead of 3 as used by Koppen, »ind 4.5 for a dry winter regime instead of 10. A map of tho climates of the western United States, based upon climatic means, drawn to test tho validity of these values, provod tliam to bo quite satisfactory. (W) ...