Precipitation Trends in the 20th Century
Aside from changes in temperature itself, perhaps the most important potential impact of future global warming will be its effect on global precipitation patterns and the frequency of severe droughts and floods. For about a century, a significant fraction of the land area of Earth has been covered by rain gauges that collect rain at individual locations and record the amount that falls over a given time period. In principle, the global network of rain gauges might be used to tell us whether noticeable changes in rainfall have already taken place during the 20th Century. Unfortunately, inferring precipitation over large areas from such point measurements is notoriously difficult. Most precipitation occurs in short bursts over small areas. This is typical of thunderstorms, which can produce torrential rains for a few minutes in one location while no rain at all falls just a few miles away. Since rain gauges are not placed in every spot, it is difficult to objectively estimate how much rain has fallen over a large area. To make matters worse, rain gauges and the way rainfall is recorded have changed over the years, and some station locations have been moved. These changes are not always well documented, making it difficult to separate real climate changes from spurious ones in the historical record.
Scientists at the Goddard Institute have compiled the available precipitation data into a form that removes these problems and used it to detect some important climatic features of 20th Century precipitation changes. Well-known climatic disturbances that have affected rainfall in the past, such as El Nino and the long-term drought in the Sahel region of Africa, are evident in this dataset. Of more interest, though, are the new things it tells us about how precipitation may respond to, and influence, global warming. We find that in much of the middle and high latitudes, precipitation has systematically increased over the 20th Century. This is consistent with changes predicted by climate models when concentrations of greenhouse gases such as carbon dioxide are assumed to increase. The dataset also helps us to put some recent damaging extreme precipitation events into a global, long-term perspective. For example, the severe 1988 drought in the midwest U.S., which raised public awareness about the possibility of global warming, was of a severity that has been repeated 24 times worldwide this century; however, the majority of the other droughts occurred in the tropics and in association with El Nino. On the other hand, the terrible midwest U.S. floods of 1993 are matched in severity by only five other events worldwide this century, and are the worst example recorded this century in the Northern Hemisphere.
Precipitation changes may not only be caused by global warming but may affect global warming as well (i.e., they provide a feedback). A peculiar feature of the warming recorded over the past century is that temperatures have warmed more during the nighttime than the daytime. This has some benefits for agriculture, because killing frosts may be less frequent, but it may adversely affect the growth of some plants and cause pests to proliferate. We find that the precipitation changes recorded over the 20th Century are well-matched to trends in cloud cover measured over the past few decades. Precipitating clouds such as nimbostratus and cumulonimbus can affect the range of daily temperature extremes by reflecting sunlight effectively and keeping the ground cool during the day, and trapping heat at their low bases at night. The observed narrowing of the daily temperature range may thus be one more piece of evidence about how Earth's hydrologic cycle is changing in response to an increasing greenhouse effect.
Dai, A., A.D. Del Genio and I.Y. Fung 1997. Clouds, precipitation and temperature range. Nature 386, 665-666.
Dai, A., I.Y. Fung and A.D. Del Genio 1997. Surface observed global land precipitation variations during 1900-1988. J. Climate 10, 2943-2962.