Special Report: Climate Change and River Flooding

4.2. North America

4.2.1. Projections

A regional model embedded in the global NCAR model revealed the existence of areas with significant increases and decreases of frequency and intensity of precipitation under 2xCO2 conditions (Mearns et al., 1994). In some areas, frequency decreases while mean daily amounts increase. In an earlier study, Mearns et al. (1990) found an increase in the day-to-day variability of precipitation.

In the GISS (Goddard Space Flight Center, Institute for Space Studies, USA) model, precipitation variability appears to change with the same sign as the mean precipitation itself, i.e. increasing in the warmer climate (3/4 of the time) (Rind et al., 1989). The hydrological cycle intensifies by about 10%.

4.2.2. Observations

Groisman and Easterling (1994a,b) reported that the North American annual precipitation time series shows an upward trend (+7.6%, significant at the 5% level) in precipitation amount during the period 1891-1990, owing mainly to increases in eastern Canada and adjacent regions of the United States. On the Century-scale the linear trends of the US and Canadian precipitation +3.9% per century (not statistically significant) and 13.7% per century (statistically significant), respectively. The increase was strongest for the last four decades (Figure 12). Especially significant were trends over the southeastern US (14% over the last 40 years) and northern Canada (17% over the last 40 years).

Iwashima et al. (1993) presented a study that shows an interdecadal increase in the frequency of maximum daily precipitation amounts over the entire USA (period: 1916-1985; 14 stations). However, Hogg, performing a similar analysis for selected Canadian stations, could not confirm this. But he did find that the 80s was the most likely decade for the largest one-day rainfall (Bruce, 1994).

4.2.3. Flooding

In North America, Lins and Michaels (1994) found statistically significant (p<0.05) increases in natural streamflow in nearly all regions of the United States. These are ascribed to an enhanced hydrologic cycle produced by an enhanced greenhouse forcing. The data set covers the years 1941-1988 and contains data from 559 gauging stations. Several criteria were used to ensure that any trends are due to actual changes in streamflow rather than to human activities or changes in measurements methods. The most significant increases were found from the Rocky Mountains eastward to the Atlantic. They conclude that these results support the hypothesis that enhanced greenhouse forcing produces an enhanced hydrological cycle.

In the United States too, there is historic evidence that a change in the climate induced changes in the flooding regimes. However, no simple proportionality exists between the scale or frequency of floods and climate variations. Knox (1993) inspected a 7000-year geological record of overbank floods for upper Mississippi river tributaries in mid-continent North America. Approximately 3300 years ago, when the climate became cooler and wetter, an abrupt shift in flood behaviour occurred, with frequent floods of a magnitude that now recurs only once every 500 years or more. Still larger floods occurred during the transition (between about AD 1250 and 1450) from the medieval warm interval to the cooler Little Ice Age. All of these changes were associated with changes in mean annual temperature of only about 1-2 K and changes in mean precipitation of 10-20%. In a similar study, Ely et al. (1993) presented a 5000-year record of extreme floods and climate change for the southwestern United States. Here, periods of especially high magnitude floods appeared to occur during transitions from cool to warm climate conditions. An analyses of modern floods shows that the "paleofloods" must have been linked to changes in large-scale atmospheric circulation patterns.

4.2.4. Analyses of recent floods

United States (Midwest), 27 June-15 August 1993
In June, Mississippi floods reached second highest level ever. Violent thunderstorms occurred in Iowa. By mid July, it was called the worst flooding in the northern Mississippi basin in over a century. The floods cover nearly 17000 square miles, one and a half time as great as the previous worst flood in this century (in 1937; 12700 square miles). Remarkably, it came in July, giving farmers less time to recover than during previous inundations, which always came in late winter or early spring.

At the beginning of the summer, an unusually strong jet stream was firmly established over the northern portion of the Mississippi basin, causing a stationary surface front where cooler drier air to the north was nearly continuously overrun by moisture-laden warmer air from the south. The result was: widespread areas of prolonged and excessive precipitation throughout large portion of the USA. The main reason is, therefore, the saturation of the soil due to heavy antecedent precipitation (van Ellen, 1995). The tentative analysis of the USA's National Meteorological Center (World Climate News, 1993) was as follows: The extreme flooding could be interpreted as being consistent with the global warming hypothesis. However, the flood cannot be considered as adding any conclusive evidence on greenhouse warming. It may well be that the ultimate "cause" of the excessive precipitation in the central USA is a combination of factors in conjunction with natural variability in the climate system. However, all these mechanisms seem less likely than a direct influence of the sea-surface- temperature anomaly in the tropical Pacific Ocean associated with ENSO. Preliminary tests using the current ENSO related sea-surface temperature anomalies in a numerical climate model at the National Meteorological Climate Centre show a response in North America that replicates the observed precipitation and temperature anomaly pattern to a considerable extent. An additional effect may be due to the draining of wetlands. These wetlands formed a buffer between the rivers and other low-lying lands, and their loss caused a concentration of water further downstream which, in turn, contributed to a larger flooding.