Identification of long term changes and evaluation of the relationships among streamflow variability and oceanic-atmospheric indices

摘要

To examine the effects of climate variability on streamflow, this thesis presents a comprehensive analysis of the streamflow variability of the continental United States and its association with oceanic-atmospheric indices. First, the presence of trends with consideration of short term and long term persistence followed by shifts over the past years in the continental U.S. streamflow were analyzed by using the non-parametric tests: Mann Kendall and Pettitt. Second, the spatio-temporal relationships between seasonal streamflow variability of continental U.S. and sea surface temperatures (SST) and 500 mbar geopotential height (Z500) of the Pacific and Atlantic were established using the singular valued decomposition (SVD) factorization technique. Finally, the SVD technique was used to determine the spatio-temporal relationships between western U.S. seasonal streamflow variability and indicators (SST and Z500) of Pacific Ocean variability.The results indicated that significant trends and shifts existed in the streamflow volumes in the past few decades in the continental U.S., and short term and long term persistence was evident in streamflows. Removal of the persistence characteristic removed the spurious trends, that could otherwise be misinterpreted. The 1970u27s were the years when major shifts took place, and most of the shift periods coincided with the phase changes of multi-decadal oscillations. A strong spatio-temporal relationship of the streamflow was established with the oceanic-atmospheric indicators and the multi-decadal oscillations (PDO, AMO) did influence these relationships. Seasonal analysis showed an improved explanation of co-variability over water-year. The lead-time approach showed that the relationships could be explained up to 9 months ahead and Z500 showed improved results over SST at shorter lead-times. Several new regions apart from the well identified index regions were identified in the oceanic region to have strong relationships with seasonal streamflows. The individual basins showed unique relationships with the indicators both spatially and temporally which could be due to their unique topographical and local factors.The contributions from this research include a better understanding of the changes in continental U.S. streamflow and the spatio-temporal relationships with oceanic-atmospheric variability that can help in better water management decisions.