Clarifying the changes of water and carbon variables in the water conservation zone of the Yellow River is important for understanding the evolution processes of water and carbon cycles and mechanism in the Yellow River Basin, ensuring water resources and ecological security, and supporting water and carbon resources management. In this study, we investigated the spatial and temporal variation characteristics of the six typical water and carbon variables (gross primary production, net primary production, carbon use efficiency, evapotranspiration, runoff, and water use efficiency) in the water conservation zone of the Yellow River and their causes at different spatial scales, using spatiotemporal statistical analysis, Geodetector, and partial correlation analysis, based on annual water and carbon variable data, meteorological data, and vegetation leaf area index data during 2001-2018. The results show that: 1) The water and carbon variables in the water conservation zone of the Yellow River and the three subregions mainly showed a fluctuating upward trend during the study period; the spatial distribution of water and carbon variables showed obvious regional heterogeneity. The interannual mean values of the carbon variables (gross primary production, net primary production, carbon use efficiency) and the water-carbon coupling variable (water use efficiency) were in the order of Zone III (Yiluo River Zone) > Zone II (Southern Weihe Mountainous Zone) > Zone I (Upper Yellow River Zone), which decreased spatially from southeast to northwest, while the water variables (evapotranspiration and runoff) generally showed a spatial distribution pattern of high in the south and northwest and low in the central area. 2) Climatic factors, especially water-heat factors, and land surface factors jointly drove the variation of the water and carbon variables in the water conservation zone of the Yellow River. The ranking of the driving forces of the water and carbon variables detected at the regional scale did not change fundamentally, and they were mainly driven by the leaf area index, temperature, and precipitation factors; the changes of carbon and water-carbon coupling variables at the subregional and grid scales were mainly driven by the leaf area index and the changes of water variables were mainly driven by net radiation and precipitation.

Hydrological element variation analysis and key parameter identification are one of the core contents of the research on hydrological processes and model simulation in the source region of the Yellow River. In order to understand the influence of freeze-thaw action on soil hydrological elements in the source region of the Yellow River, this study compared the data from the MQ seasonally frozen soil observatory and the KQ permafrost observatory. Based on the data of field monitoring, variability of soil hydrological elements was examined and simulated by the freeze-thaw module of the HYDRUS-1D software. The results showed that: 1) The freeze-thaw process of soil leads to changes of soil hydrological elements. The saturated soil moisture (θ_s) of seasonally frozen soil is higher during cold period than warm period, the residual soil moisture (θ_r) is the opposite. For permafrost, θ_s during cold period is higher than warm period, θ_r during cold period is lower than warm period in the deep layer only. 2) The soil water content of both seasonally frozen soil and permafrost shows a U-shaped change during freezing within the year, and the water storage of seasonally frozen soil decreases faster than permafrost during the rapid freezing period. 3) The bottom leakage flux of seasonally frozen soil and permafrost both decreases continuously during the stable frozen period, while the bottom leakage flux of permafrost decreases to 0 because of the baseplate of permafrost active layer, and remains unchanged. 4) The frozen peak point of both seasonally frozen soil and permafrost exists at 20 cm below soil surface. Soil water flows bilaterally to the frozen peak point during the rapid freezing period, and flows bilaterally out from the frozen peak point during the rapid thawing period. This study has important theoretical and practical value for understanding of the hydrological process of frozen soil more profoundly and optimizing the management of water resources in the source region of the Yellow River.

The calculation of the physical quantity of aquatic ecological products is the basis for the accounting and realization of the value of aquatic ecological products. In order to objectively reflect the relationship between the value of aquatic ecological products and water cycle processes, this study took the Huangshui River Basin in the upper reaches of the Yellow River as an example, and calculated the physical quantity and value of aquatic ecological products (water resources supply, water resources stock, soil conservation, flood regulation and storage, and water quality purification) in the basin from 1986 to 2015 by means of multi-process simulation of the water cycle in the basin, scenario analysis, and socio-economic condition investigation. The value composition and spatiotemporal distribution pattern of aquatic ecological products were systematically analyzed. The results show that the total value of aquatic ecological products in the Huangshui River Basin from 1986 to 1995 was 60.481 billion yuan/a, which increased by 10.87% and 24.49% in 1996-2005 and 2005-2015, respectively. The contribution of water resources storage was the highest (66.59%-69.53%), followed by reservoir flood control, water supply, and water purification (accounting for 18.91%-23.54%, 5.08%-5.53%, and 4.77%-6.12%, respectively), and the contribution of soil conservation was negligible (0.02%). The value of aquatic ecological products in sub-basins usually decreased from the upstream to the downstream, but the values of aquatic ecological products have increased significantly in sub-basins with water conservancy projects. This study provides a theoretical and technical support for the management of water ecosystems in the upper reaches of the Yellow River.

The water conservation zone of the Yellow River is a critical ecological function zone in China and further study is needed to maintain its ecosystem health and promote high-quality development. This study employed various analytical methods such as trend analysis, LOWESS (locally weighted scatterplot smoothing), Pearson and partial correlation analyses, and residual analysis, to investigate the spatiotemporal variation of vegetation and its driving factors in the Yellow River water conservation zone from 1982 to 2015. It also divided the conservation zone into plain and plateau regions based on topographical differences to examine the intra-regional disparities. The results reveal that: 1) The growing season normalized difference vegetation index (NDVI) in the conservation zone exhibited a significant growth trend, mainly driven by the rapid growth of NDVI in spring and autumn, with varying growth rates at different regional scales (0.0009/a for the whole area, 0.0007/a for the plateau area, and 0.0016/a for the plain area). 2) The correlation between regional NDVI and precipitation and temperature differed through time and across space. During the growing season, NDVI was negatively and positively correlated with precipitation and temperature, respectively. In spring and summer, temperature was the primary controlling factor, whereas in autumn, both precipitation and temperature were involved. Except for temperature in spring, the plain area was controlled by precipitation and temperature in other periods. The growing season NDVI of the conservation zone had a positive correlation with rainfall and temperature in 50.4% and 91.4% of cases, respectively. 3) The NDVI variation trend went through some stages—it stagnated in 1995-2015, during which the correlation between NDVI and rainfall shifted from negative to positive, but before 1995 and after 2015, NDVI increased significantly and had negative and positive correlations with rainfall and temperature, respectively. 4) Human activities and climate change jointly were the primary drivers of NDVI growth in the region, accounting for 74.30% of the total area. The second largest factor of vegetation growth was climate change, which accounted for 11.48% of the area, and vegetation decline caused by human activities accounted for 6.23% of the area. These findings suggest that ecological engineering construction effectively promoted vegetation recovery; however, human activities such as urbanization can also disrupt vegetation growth.