Macroscale composite principal-monotonicity distributed projection of climate-induced hydrologic changes: Tempo-spatial variabilities, and variation decompositions

Abstract

A macroscale composite principal-monotonicity distributed hydro-projection system is developed through hybridizing advanced statistical hydroclimatic systems analysis methods of verified superiorities. The system can enhance accuracies and feasibilities of climate-informed hydrologic projections, quantitatively reveal climate-induced hydro-magnitude and -variability changes, and decompose variations of the changes with factors such as time, space, scales and greenhouse gas (GHG) emissions. The system of high adaptability for worldwide watersheds is applied to a representative large cold-region watershed, Athabasca River Basin in Canada. Incremental ratios of river discharges with climate change are projected to enlarge at fine timescales, e.g., from 3.4% (octo-decadal) to 7.0% (monthly). According to mixed-level multifactorial analyses, variations of the ratios are dominated by catchments (59.2%) and GHG emissions (19.6%) at annual and longer timescales, and the dominance increases with timescales. At intra-annual timescales, the dominator becomes interactive tempo-spatial variations (62.6%) of the ratios. GHG emissions and their interaction with catchments pose significant or even leading effects on trends of discharge changes. Ratios of changes (i.e., decreases mostly) in tempo-spatial variabilities of discharges are generally the most significant for interannual variabilities, in warm seasons, at downstream catchments, under high GHG emissions, and by the late 21st century. Their variation is dominated by individual factorial effects (78.0%), e.g., catchments (50.3%) and GHG emissions (16.2%); the dominance declines from interannual (89.0%), temporal (79.4%), intra-annual (75.6%) to spatial (67.9%) variabilities. Interactions of the factors in explaining variations of spatial and interannual variabilities are non-negligible in warm and cool seasons, respectively. These findings can help advance macroscale distributed climate-informed hydrologic projections and, accordingly, scientize assessment, adaptation and mitigation of hydrologic hazards over climate-sensitive cold regions.