乌江流域蓝绿水对气候和土地利用变化的响应

为了更深入地研究气候变化和土地利用变化对乌江流域蓝绿水时空分布的影响，该研究采用SWAT（soil and water assessment tool）模型，利用1985—2019年逐日气象数据以及1990年、2000年和2010年土地利用数据，设定多种变化情景，定量分析了蓝水绿水对乌江流域气候和土地利用变化的响应。同时，还利用5个CMIP6（Coupled Model Intercomparison Project Phase 6）模式集合平均（multi-modal ensemble，MME）以及FLUS（future land-use simulation）模型，预估了乌江流域在2015—2100年间蓝绿水的时空变化。结果表明：1）SWAT模型在乌江流域的校准期和验证期的决定系数（coefficient of determination，R²）分别为0.94、0.87，纳什系数（nash-sutcliffe coefficient，NSE）分别为0.94、0.77，FLUS模型的Kappa系数为0.764，证明其在流域水文模拟和未来土地利用预估方面的可靠性。2）在1985—2014年间，乌江流域的蓝水、绿水都呈现出先增加后减少的趋势。蓝水、绿水均先分别增加了137.3、7.9 mm/a，然后分别减少了127.5、12.7 mm/a。气候变化对蓝水变化的影响贡献占比为99%，对绿水变化的影响贡献占比平均为88%，表明气候变化在蓝水和绿水变化中的主导地位。3）到2040年，城镇用地面积几乎比2010年扩张了近1.5倍。在2015—2100年，无论是在SSP2-4.5还是SSP5-8.5情景下，蓝水、绿水都呈增加趋势。因此未来需要合理规划乌江流域的土地利用，控制城镇化速度。研究结果可为乌江流域合理规划土地利用、实现水资源综合管理提供支持和参考。

Response of blue and green water to climate and land use changes: A study in the Wujiang River Basin, China

Blue and green water resources are essential for human livelihood and the sustainable development of ecosystems. It is crucial to thoroughly examine the effects of climate change and land use change on these resources. This study takes the Wujiang River Basin as an example, which, as a typical mountainous basin, has more prominent water resource problems. Firstly, two sets of land use data from 1990, 2000 and 2010 were selected, considering various climate and land use change scenarios. The impact of climate and land use changes on blue and green water resources in the Wujiang River Basin from 1985 to 2014 was then quantitatively analyzed using the soil and water assessment tool (SWAT). Subsequently, the future land-use simulation (FLUS) model was employed to simulate land use in 2040, based on two periods of land use data (2000 and 2020). Finally, the SWAT model, driven by future land use data and two scenarios from the CMIP6 model (SSP2-4.5 and SSP5-8.5), projected the blue and green water resources in the Wujiang River Basin for the period 2015-2100. The results indicate the following: 1) The SWAT model demonstrated good performance in simulating the water cycle process in the Wujiang River Basin, with nash-sutcliffe coefficient (NSE) and coefficient of determination (R²) values of 0.94 and 0.94 for the calibration period, and 0.77 and 0.87 for the validation period, respectively. 2) The FLUS model effectively simulated land use in 2020, achieving a Kappa coefficient of 0.764 and demonstrating its applicability for future land use simulations. 3) From 1985 to 2014, land use changes revealed reversibility patterns, with agricultural land being converted mainly into forest land and grassland, forest land being converted mainly into agricultural land and grassland, and grassland being converted mainly into agricultural land and forest land. 4) The period from 1985 to 2014 witnessed an initial increase followed by a decrease in blue and green water resources in the Wujiang River Basin, with blue water increasing by 137.3 mm/a and then decreasing by 127.5 mm/a, and green water increasing by 7.9 mm/a and then decreasing by 12.7 mm/a. Climate change accounts for 99% of blue water change and 85% to 90% of green water change, highlighting its predominant role in these shifts. 5) Urban land is projected to expand almost 1.5 times in 2040 compared to 2010. Both scenarios indicate an upward trend in blue and green water resources when contrasted with the historical base period (1985-2014), with the SSP5-8.5 forward period displaying the most substantial increase. This trend amplifies the risk of flooding and drought occurrences. These findings offer valuable insights for informed land use planning, integrated water resources management, and decision-making concerning future water resource management in the Wujiang River Basin.