To estimate the importance of climatic changes to society, researchers need not, however, study global average temperature so much as the possible regional distribution of evolving patterns of climatic change in the future. Will it, in the year 2010, be drier in Iowa, wetter in Africa, more humid in New York, or too hot in India? Unfortunately, to predict reliably the fine-scale regional response of variables, such as temperature and rainfall, requires climatic models of greater complexity (and cost) than are currently available. At present, there is simply no consensus among knowledgeable atmospheric scientists that the regional predictions of state-of-the-art models are reliable. Nevertheless, most experts agree that the following coherent regional features might well occur by about the year 2035: wetter subtropical monsoonal rain belts; longer growing seasons in high latitudes; wetter springs in high and middle latitudes; drier midsummer conditions in some mid-latitude areas (a potentially serious agricultural and water supply problem in major grain-producing nations); increased probability of extreme heat waves (with possible health consequences for people and animals in already warm climates); and an increase in sea level by a few tens of centimetres. Considerable uncertainty remains in these regional estimates, even though many plausible scenarios have been investigated.