DETECTION OF CHANGES IN STREAMFLOW AND FLOODS RESULTING FROM CLIMATE FLUCTUATIONS AND LAND USE-DRAINAGE CHANGES

Detection of effects of changing climate on the hydrologic responses o f rivers can be further complicated by changes in land use, drainage, and water use. To discern effects of human-caused changes in a basin a nd those due to precipitation over time, a comparison was made of annu al mean flows and peakflows in Midwestern basins that experienced incr eases in annual precipitation and heavy rain events during 1940-1990. Two pairs of basins, one pair in a rural area and one pair in an urban ized area, were selected for in-pair comparisons, with one basin in ea ch pair experiencing more land use and drainage changes during 1940-19 90 than the other basin. All basins experienced significant upward tre nds in annual precipitation and annual mean flows. Human-produced chan ges affecting runoff in both rural basins accounted for about two-thir ds of the fluctuations in the mean Bows, and precipitation changes acc ounted for the other third. However, much of the change in peakflows i n the rural basin undergoing sizable changes in drainage was due to th ese changes (85%) versus 75% in the rural basin without comparable shi fts in drainage. The mean and peak flows of the two urban basins showe d considerably more response to precipitation shifts than those of the two rural basins. The urbanized area doubled within one urban basin d uring 1940-1990, and these land use changes explained much more of the increase in mean flows and peakflows there than in the urban basin wi th less change in land use. By 1990 precipitation accounted for 69% of the upward trend in mean flows since 1941 in the heavily developed ur ban basin, as compared to 37% of the trend in the less settled urban b asin. For purposes of assessing climate change, the precipitation chan ges over fifty years in all basins produced marked uptrends in basin s treamflow, but the magnitude of the precipitation effect was masked by the land use and drainage changes. The results illustrate the need fo r careful analysis of natural basin characteristics (soils and basin s hape), land use and drainage changes, and of various precipitation con ditions if the influence of shifting precipitation on hydrologic condi tions is to be detected, accurately measured, and correctly interprete d. For such studies the paired basin comparison techniques appears to be a valuable approach.