Cumulative ecosystem response to Hydraulic Engineering Infrastructure Projects in an arid basin

Hydraulic Engineering Infrastructure Projects (HEIPs) typically show profound effects on hydrological systems and ecosystems. However, data restrictions have limited the exploration of the influences of compound HEIPs on ecosystems to a few studies. This study proposes a watershed-wide ecosystem assessment framework to investigate the impact of HEIPs in the Tarim River Headwaters-Hotan River Basin on the ecosystem of the arid zone. The framework includes a deep learning-meta cellular automata algorithm (DLMCAA) based on the spatiotemporal characteristics of HEIPs and hydro-meteorological and human activities. Moreover, the spatiotemporal relationships between compound HEIPs and ecosystem variances were quantified. The framework including DLMCAA showed a good performance in simulating landcover in 2020, with a Kappa coefficient of 0.89. Therefore, the DLMCAA could be used to simulate and predict ecosystem changes under the HEIPs, which suggested that the framework is effective and practical. An analysis of the spatiotemporal distribution of each ecosystem from 1980 to 2020 showed that the low shrub ecosystems changed most significantly (26.38 %) between 1980 and 2020. Also, the use of spatially driven hydrological project data from different ABC scenarios showed that ecosystems driven by HEIPs were more stable compared to those without HEIPs under future climate change. In particular, the DLMCAA indicated that compound HEIPs had a more positive impact on ecosystem oases in arid lands compared with that of single HEIPs. The results of this study can serve as a scientific reference for assessing the impact of HEIPs, as well as for understanding ecosystem changes and facilitating sustainable water resource management in the arid regions.