气候变化领域集成服务门户

The present study focuses on change detection and attribution analysis for a case study of Allegheny river catchment (at two unregulated sites namely Eldred and Salamanca) inUSA. The proposed methodology involves, multiple change-point detection (MCPD) techniques i. e., Binary Segmentation based cumulative sum algorithms with Monte Carlo based threshold (BSCSth) and Bayesian Information Criteria (BIC) based penalty (BSCSBIC) and validating the two techniques using standard performance measures for different hypothetical riverflow time-series. The proposed BSCSth techniquewas applied to the Allegheny river at two sites for the series of different hydrological alteration indicators to identify the location of change-point in mean, and the hydrological regime shifts. To perform attribution analysis, hydrologic simulations were carried out using Sacramento model for the identified segments. The overall hydrologic alterations of selected annual flow metrics extracted from daily simulated flows and observed flow values were also estimated. The results of MCPD analysis showed that overall three hydrological regimes comprising of three segments i. e., near natural period (NNP, 1940-1955), low impact period (LIP, 1956-1966) and high impact period (HIP, 1967-2014) were identified for both the sites. The results of attribution analysis for the three cases (case-1: NNP & LIP; case-2: LIP & HIP; case-3: NNP & HIP) showed that for the case-1, the changes of flow regimes in Eldred and Salamanca were predominantly affected by climate-induced processes. But, for case-2 and case-3, it was seen that there were magnification of human-induced changes probably due to the landuse change (forest and rangeland increased by 18.31% (19.79%) for Eldred (Salamanca) from 1940 to 1992), construction of Kinzua dam (in the year 1966) and other anthropogenic pressures. Thus, the results of the study can be helpful for quantifying the main drivers of hydrological changes in river basin, and planning suitable water resource management strategies in the basin. (c) 2019 Elsevier B.V. All rights reserved.