System devised in 1918 by Wladimir Peter Köppen (1846–1940), with modifications that were completed in 1936, by which climates are divided into six broad groups according to the major vegetation types associated with them, broadly determined by critical temperature and the seasonality of precipitation. For example, a summer temperature of 10  °C defines the poleward limit of tree growth; a winter temperature of 18  °C is critical for certain tropical plants; a temperature of −3  °C indicates some period of regular snow cover. The groups are: (A) tropical rainy climates with temperatures in the coldest month higher than 18  °C; (B) arid climates; (C) warm, temperate, rainy climates in which temperatures in the coldest month are between 3  °C and 18  ° C, and in the warmest month higher than 10  °C; (D) rainy climates typical of boreal forest, in which temperatures in the coldest month are lower than −3 °C (in US usage modified to 0  °C), and in the warmest month higher than 10  °C; (E) tundra, in which temperatures in the warmest month are 0–10  °C; (F) permanent frost and ice caps, in which temperatures in the warmest month are below 0  °C. Subsets of the main classes (written as capital and lower case letters, e.g. Cs) are: absence of a dry season (f); a dry summer season (s); a dry winter season (w); a monsoon climate, with a dry season and rains at other times (m). Arid climates (B) are subdivided into semi-arid steppe-type (S), and arid desert (W). The temperatures within class (B) are indicated as: mean annual temperature higher than 18  °C (h); mean annual temperature lower than 18 °C with the warmest month higher than 18  °C (k); mean annual temperature and warmest month both lower than 18  °C (k′). Some criticism of the system centres on its arbitrary criteria of temperature associated with fixed boundaries, and the failure of data on temperature and precipitation to account fully for the effectiveness of the precipitation. See also strahler climate classification; thornthwaite climate classification.