Distributional change of monthly precipitation due to climate change: comprehensive examination of dataset in southeastern United States |
| |
| Authors: | Hui Wang Rebecca Killick Xiang Fu |
| |
| Affiliation: | 1. Bureau of Economic Geology, School of Geoscience, The University of Texas at Austin, Austin, TX, USA;2. Department of Mathematics and Statistics, Lancaster University, Bailrigg, Lancaster, UK;3. State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan, China |
| |
| Abstract: | A number of watersheds are selected from the Hydro‐Climate Data Network over southeastern United States to examine possible changes in hydrological time series, e.g. precipitation, introduced by changing climate. Possible changes in monthly precipitation are examined by three different methods to detect second order stationarity, abrupt changes in the variance and smooth changes in quantiles of the time series. An analysis of second order stationarity shows that precipitation in eight of the 56 watersheds display nonstationary behaviour. Change‐point analyses reveal that changes in the long‐term variance of monthly precipitation are only detected for a few sites. As a complementary analysis tool, quantile regression aims to detect potential changes of different percentiles of the monthly precipitation over time. Several sites show diverging trends in the quantiles, which implies that the range and thus variance of the data, is increasing. As distinct change‐points are not identified, this suggests that the effect is small and cumulative. Results are analysed in detail, and possible explanations are provided. This type of thorough analysis provides a basis for understanding the possible redistribution of water cycle. It also provides implications for water resources management and hydrological engineering facility design and planning. Copyright © 2013 John Wiley & Sons, Ltd. |
| |
| Keywords: | climate change climate variability change‐point analysis quantile regression hydrological time series |
|
|
