The Mekong river is one of the most complex river systems in the world that is shared by six nations in Southeast Asia. The river still remains relatively undammed (most existing dams are in the tributaries and are small), and its hydrology today is dominated by large natural flow variations that support the highly productive agricultural and riverine ecological systems; however, this is changing due to the alterations in land use and construction of new dams both in the tributaries the mainstream. Understanding the changes in surface water dynamics is therefore crucial to provide realistic future predictions of changes in downstream floodplain and riverine ecology due to the construction of dams in the upstream. While the existing dams have caused little impact on mainstream flows, those under construction and planned are likely to cause severe and potentially permanent damage to downstream hydro-agro-ecological systems, and adversely impact the livelihood of millions. Here, using hydrodynamic model simulations (CaMa-Flood), we show that the effects of flow regulation on downstream river-floodplain dynamics are relatively predictable along the mainstream Mekong, but flow regulations could potentially disrupt the flood dynamics in the Tonle Sap River (TSR) and small distributaries in the Mekong Delta. Results suggest that TSR flow reversal could cease if the Mekong flood pulse is dampened by 50% and delayed by one-month. While flood occurrence in the vicinity of the Tonle Sap Lake and middle reach of the delta could increase due to enhanced low flow, it could decrease by up to five months in other areas due to dampened high flow, particularly during dry years. Further, areas flooded for less than five months and over six months are likely to be impacted significantly by flow regulations, but those flooded for 5-6 months could be impacted the least.

Future socio-economic and climate changes can profoundly impact water resources, food production, bioenergy generation, and land use, leading to a broad range of societal problems. In this study, we performed future projections by using a land integrated model, MIROC-INTEG-LAND, that considers land surface physics, ecosystems, water management, crop growth, and land use, under various socio-economic scenarios (Shared Socio-economic Pathways, SSPs). Under the sustainability scenario (SSP1), demands for food and bioenergy are kept low, so that the increase in cropland areas for food and bioenergy are suppressed. On the contrary, in the middle of the road and regional rivalry scenarios (SSP2 and SSP3), cropland areas are projected to increase due to high demand for food and bioenergy. The expansion of cropland areas is projected to increase the water demand for irrigation and CO2 emissions due to land use change. MIROC-INTEG-LAND simulations indicate that the impacts of the CO2 fertilization effect and climate change on crop yields are comparable, with the latter being greater than the former under climate scenarios with high greenhouse gas concentrations. We also show that the CO2 fertilization effects and climate change play important roles in changes in food cropland area, water demand for irrigation, and CO2 emissions due to land use change. Our results underscore the importance of considering Earth-human system interactions when developing future socio-economic scenarios and studying climate change impacts.

Irrigation parameterizations in land surface models have been advanced over the past decade, but the newly available data from the Soil Moisture Active Passive (SMAP) satellite has seldom been used to improve irrigation modeling. Here, we investigate the potential of assimilating SMAP soil moisture (SM) data into the Community Land Model (CLM) to improve irrigation representation. Simulations are conducted at 3 arc-minute resolution over the highly irrigated region in the central US, fully enclosing the upstream areas of the river basins draining over the High Plains Aquifer (i.e., the Missouri and Arkansas), and Colorado River basins. We test the original CLM4.5 irrigation scheme and two new irrigation parameterizations using SMAP data assimilation by: (1) directly integrating raw SMAP data, and (2) integrating SMAP data using 1-D Kalman Filter (KF) smoother. An a priori scaling approach is also used to account for bias correction of the shortly-recorded SMAP data based on the ground observations, enabling us to use SMAP for out-of-sample tests (i.e., assessment of the new parameterizations during a non-SMAP period). The ground-based SM observations from three monitoring networks, namely Soil Climate Analysis Network (SCAN), US Climate Reference Network (USCRN), and SNOwpack TELemetry (SNOTEL) are employed for bias correcting SMAP data and validating SM simulations. Results show that SMAP data assimilation using 1-D KF significantly improves irrigation simulations. Bias correction of SMAP data further improves results from KF assimilation in some regions. However, the improvements are small compared to those achieved from 1-D KF application alone, indicating the robustness of using SMAP data and KF globally even for the regions where ground-based data are not available for bias correction. The data assimilation also improves the accuracy of the temporal dynamics and vertical profile of simulated SM. These results are expected to provide a basis for improved modeling of irrigation water use and land-atmosphere interactions.