POTENTIAL IMPACT OF CLIMATE CHANGE ON LONG-TERM STREAM DISCHARGE IN NEW YORK STATE
To determine whether precipitation and/or discharge data exhibit long-term trend, annual precipitation and annual mean discharge data were subjected to Mann-Kendall test. Though statistically significant trends were detected, because precipitation and discharge data are marked by strong persistence, it is likely that the statistical significance is affected by the presence of persistence. Accordingly, we performed the Mann-Kendall test after pre-whitening both precipitation and discharge data. The pre-whitening process involved three steps: (i) removing the trend using the Thiel-Sen estimate of the slope, (ii) removing the effect of lag-1 autocorrelation from the de-trended data and (iii) adding the trend back to the data from step (ii).
Results do not show a consistent pattern in the five watersheds with respect to stream discharge. Furthermore, patterns in discharge trends do not always correlate with precipitation trends. Further evaluation of the stream discharge trends is looking at changes in land-use, its impact on stream discharge and potential influence on the observed trends in the data.