Streamflow trends in Indiana; effects of long term persistence, precipitation and subsurface drains
Dates
Year
2008
Citation
Kumar, Sanjiv, Merwade, Venkatesh, Kam, Jonghun, and Thurner, Kensey, 2008, Streamflow trends in Indiana; effects of long term persistence, precipitation and subsurface drains: Journal of Hydrology , v. 374, no. 1/2, p. 171-183.
Summary
Streamflow trends in Indiana are evaluated for 31 USGS gauging stations that have 50 years or more continuous unregulated streamflow records. Trends are computed for selected streamflow statistics to include annual and seasonal low, medium, and high flow conditions by using four variations of the Mann-Kendall test. These variations include: (i) Mann-Kendall without autocorrelation, (ii) Mann-Kendall with lag-1 autocorrelation and trend-free pre-whitening, (iii) Mann-Kendall with complete autocorrelation structure, and (iv) Mann-Kendall with long term persistence. Mann-Kendall test is also applied to the precipitation data using the above four approaches to explore the relationship between precipitation and stream flow trends. Overall, [...]
Summary
Streamflow trends in Indiana are evaluated for 31 USGS gauging stations that have 50 years or more continuous unregulated streamflow records. Trends are computed for selected streamflow statistics to include annual and seasonal low, medium, and high flow conditions by using four variations of the Mann-Kendall test. These variations include: (i) Mann-Kendall without autocorrelation, (ii) Mann-Kendall with lag-1 autocorrelation and trend-free pre-whitening, (iii) Mann-Kendall with complete autocorrelation structure, and (iv) Mann-Kendall with long term persistence. Mann-Kendall test is also applied to the precipitation data using the above four approaches to explore the relationship between precipitation and stream flow trends. Overall, there is an increasing trend in low and medium flow conditions across Indiana. However, these trend results are affected by the selection of flow statistic, annual/seasonal statistics, data length and consideration of autocorrelation structure. Specifically, the number of stations that show trends in low and medium flow statistics decreases significantly (up to 70%) when autocorrelation structure is considered in the analysis. The precipitation trends, however, are not affected by the autocorrelation structure, suggesting that factors other than precipitation trends are related to streamflow trends in Indiana. In addition, high flows are found to be less correlated to precipitation compared to low flows in heavily drained northern Indiana. This suggests that the subsurface drainage is playing a role in the observed streamflow trends in Indiana, but this issue needs further investigation. Overall, this study provides an elaborate view of past streamflow trends in Indiana which should be useful for further research including understanding the impact of future biofuel activities on Indiana hydrology and climate.
