In 1931 Thornthwaite devised a complex and empirical classification which is very close to Koppen’s scheme. It also attempts to define climatic boundaries quantitatively and is based on plant associations. Like Koppen’s this classification also employs combination of letter symbols to designate the climatic types, its subdivisions and other groups.
However, Thornthwaite’s classification is based on precipitation effectiveness and thermal efficiency (temperature efficiency). Under this classification climatic types were subdivided by the use of a term to denote the seasonal distribution of precipitation.
The climatic types and their boundaries were defined empirically by observing the characteristics of natural vegetation, soil, and the drainage pattern.
Thornthwaite established the fact that not only the amount of precipitation, but the rate of evaporation as well is significant for the growth of natural vegetation.
Thus, besides the precipitation amount and the evaporation rate, temperature was made a very important basis for Thornthwaite’s climatic classification. An expression for precipitation efficiency was obtained by relating measurements of pan evaporation to temperature and precipitation.
For each month the ratio 11.5 (rt-10)10/9, where r = mean monthly rainfall (in inches) t = mean monthly temperature (in °F) is calculated. The sum of the 12 monthly ratios gives the precipitation effectiveness (also called precipitation efficiency) index.
In other words, the effectiveness of precipitation is taken to be a function of precipitation and evaporation, and is calculated by dividing the monthly precipitation by the monthly evaporation to get the P/E ratio.
On the basis of P/E indices and boundary values for the major vegetation regions, five humidity provinces were defined.
T/E Index = sum of 12 monthly values of (T-32)/4, where T is the mean monthly temperature in °F.
On the basis of the seasonal distribution of precipitation the humidity provinces were subdivided into the following:
r-Rainfall adequate in all seasons.
s-Rainfall deficient in summer.
w-Rainfall deficient in winter.
When precipitation effectiveness, seasonal distribution of rainfall and thermal efficiency are taken together, there would be in all 120 climatic types, at least on theoretical grounds. However, Thornthwaite has shown only 32 climatic types on the world map depicting his 1931 climatic classification (Table 38.5).
This classification is almost similar to Koppen’s classification in so far as it defines climatic boundaries quantitatively, and it is also based on vegetation. Like Koppen’s scheme it also makes use of letter combinations to designate climatic types.
However, it differs from Koppen’s classification on two scores: first, he introduced an expression for precipitation efficiency, and second, he made use of an index of thermal efficiency. Thornthwaite makes moisture the primary classificatory factor for T/E index of over 31 (the taiga/cool temperate boundary).
Since Thornthwaite adopted the precipitation effectiveness and temperature efficiency indices for his climatic classification, the delimitation of the climatic boundaries becomes difficult and vague.
Besides, under this classification, climatic types are almost three times greater than Koppen’s climatic types. But the point to be remembered is that in Thornthwaite’s classification the letter symbols used are relatively less in number than those in Koppen’s classification.
Another characteristic feature is that the climatic types in this classification can be recognised by the letter symbols alone. In this scheme of climatic classification the lack of adequate climatic data presented a serious handicap.
In this context it may be mentioned that since some of the weather stations do not keep a continuous record of the rate of evaporation, there was a great difficulty in procuring the records thereof.
Like Koppen’s classification, this classification too does not permit comparison form one locality with another. This is because of the fact that monthly temperature and precipitation permit estimates of the important factor of precipitation effectiveness, but do not measure it with enough precision.
Lastly, it may be pointed out that the climatic classifications as devised by Koppen and Thornthwaite are more useful and appealing to zoologists, botanists, and geographers.
But these schemes of classification are not so useful for meteorologists and climatologists because the interplay between the weather elements and other climatic factors is not clearly shown.