Using the multi-model ensemble projections of wheat and rice in India during the 21st century from the Inter-Sectoral Impact Model Intercomparison Project, this study assessed the future changes in crop yield under the RCP 8.5 emission scenario. The results show that the multi-model ensemble simulations generally reproduce the spatial variability in crop yield that is represented by ground measurements. Furthermore, the simulations reproduce the response of crop yield to climate changes, which is characterized by negative correlations between crop yield and temperature and positive correlations between crop yield and precipitation. Under the RCP8.5 emission scenario, temperature and precipitation during the growing season of wheat and rice will increase. Generally, temperature will increase at a higher rate than precipitation; the increases in temperature and precipitation during the wheat growing season will be stronger than that during the rice growing season. In terms of spatial dimension, the increase in temperature will be weakened gradually from the north to the south, while that of precipitation will be intensified gradually from the north to the south. Temperature increase in the wheat producing areas will be stronger than that in the wheat non-producing area, while precipitation increase in the wheat producing areas will be weaker than that in the wheat non-producing areas. However, the scenario is predicted to be reversed for the rice producing areas. In response to the climate changes, wheat and rice yields will decrease in the 21st century, particularly in the second half. The decrease in wheat yield will be greater than that of rice yield. In the first half of the 21st century, wheat and rice yields are predicted to decrease at the rates of 1.3%/10 a (P < 0.001) and 0.7%/10 a (P < 0.05), respectively. In the second half of the 21st century, they are predicted to decrease at the rates of 4.9%/10 a (P < 0.001) and 4.4%/10 a (P < 0.05), respectively. The drought stress resulting from climate warming might be the main reason for this projected yield reduction. The greatest decrease in wheat yield (as high as 60%) will occur in the southwest part of the Deccan Plateau, and the greatest decrease in rice yield (as high as 50%) will occur in the northern part of the Gangetic Plain. These findings suggest that food supply in India will face extreme challenges under the future climate change scenarios.

The Qinling Mountains serve as a dividing line for China's climate. The dearth of precipitation data in high-altitude areas limits the study of climate and water resource changes in this region. To determine the spatial distribution of precipitation, the weather research forecasting (WRF) model was used to simulate summer precipitation in the Qinling Mountains and surrounding areas in 2017 by employing three convective parameterization schemes (KF, BMJ and GF). Both the simulation results and satellite data reveal that precipitation is consistent with the terrain's topography, and there is a spatial pattern of strip distribution from south to north accompanied by a high precipitation zone in the Qinling Mountains. The mountains have an evident blocking effect on the northward transport of water vapour, resulting in precipitation on the southern slope of the Qinling Mountains being significantly greater than that on the northern slope. However, the analogue value is larger than the satellite precipitation data. The primary reason that more precipitation is observed in the KF scheme simulation is that the strong convective instability leads to an overestimation of convective precipitation. The GF scheme simulates grid-scale precipitation, which result in more large-scale precipitation due to atmospheric wetness. The simulated values of the BMJ scheme are closest to the observed values. Increasing the resolution to 2 km can significantly improve the simulation level of the model in the Qinling Mountains, and there is great potential for improving precipitation estimation at the highest elevations of the mountains by using a suitable cumulus parameterization scheme.

Daily temperature data at 21 stations across the Pearl River Delta during the period of 1967-2015 were selected, and these stations were divided into urban stations and suburban stations based on population, population density and DMSP/OLS data. Meanwhile, spatio-temporal variations due to urbanization were investigated by comparing the temperature changes at urban and suburban stations. The results show that: (1) In the past 50 years, the annual average temperature, the average maximum temperature and the average minimum temperature in the Pearl River Delta region increased significantly, and the average minimum temperature increased at the highest rate, or 1.05-1.16 times of the average temperature and 0.95-1.32 times of the average maximum temperature, respectively. Among them, the seasonal difference of annual average temperature change rate is generally manifested as the strongest temperature increase in autumn and winter, with the temperature increase rate higher than 0.3 ℃/10a, and the temperature increase rate is weaker in spring and summer, with the temperature increase rate as low as 0.16 ℃/10a. (2) Using urban and Sea Surface Temperature (SST) comparison to study the urbanization effect, affected by urbanization, the annual average temperature of the Pearl River Delta is increasing by 0.096 ℃/10a. (3) Using urban and rural comparison to study the urbanization effect, from 1967 to 2015, urbanization leads to the temperature increase in urban areas. Generally speaking, urbanization contributes the most to the increase of average minimum temperature. Meanwhile, the seasonal difference in the contribution rate of urbanization to the annual average temperature change indicates that the increase rate is stronger in summer and winter, with the contribution rate higher than 11.8%, and lower in spring and autumn, with the contribution rate being only 4.46%. (3) The selection of stations division method, the different periods of urbanization development and time scale of research all lead to the uncertainty of the research results on the warming effect of urbanization. Different stations classification methods indicate that urbanization contributes the most to the increase of the minimum temperature, and for a longer time scale, the contribution rate of urbanization to the increase of the minimum temperature is up to 38.6%.

A new method was proposed to identify drought-flood transition events by combining a drought-flood index (Standard Weighted Average Precipitation, SWAP) with the multi-threshold theory. This method was tested in the Yangtze River Basin using daily precipitation data from 212 stations for the 1961-2017 period. With this method, the meteorological drought and rainstorm flood in the representative station were identified, and representative regional drought-flood transition events and spatiotemporal patterns of drought-flood transition were analyzed. The results show that: SWAP is an effective index to identify the meteorological drought and rainstorm flood. K-means clustering can classify similar drought-flood transition events into one cluster. The drought event plays a dominate role in drought-flood transition events for the middle and lower reaches of the Yangtze River Basin in 2011, and the drought lasts for a much longer duration than the flood during the drought-flood transition event. Drought-flood transition events show an obvious regional pattern for the Yangtze River Basin with low frequency for the upper reaches and high frequency for the middle and lower reaches. In addition, the drought-flood transition frequency presents an increasing trend recently for most parts of the Yangtze River Basin. Overall, the results imply that the proposed method combining meteorological drought index with multi-threshold theory is capable of identifying drought-flood transition events, and can be further used for predicting drought-flood transition events.

Environmental change and human-land relationship evolution over the Qinghai-Tibet Plateau are the frontiers of global change science. The Lajia Ruins in the Guanting Basin situated in the northeast margin of the Qinghai-Tibet Plateau has preserved many records of the Holocene environmental change, human activities, complex surface process evolution and abrupt catastrophic events. Therefore, the Lajia Ruins has received extensive attention from scientists from different fields, and has also attracted attention from world media. It is still quite controversial about the causation and age of the prehistoric catastrophe. After extensive field works over the years, a complete pedo-stratigraphy profile of soil and sediments has been found in the northern part of the Lajia Ruins. On the basis of field pedo-sedimentological observation and description, as well as physic-chemical analysis and OSL dating on the soil and sediment samples taken from the profile, a pedo-stratigraphic sequence and related chronology have been established in the Lajia Ruins. The results show that the profile has been developed from 12000 a BP. The pedo-stratigraphic sequence includes the Late Pleistocene Malan loess (L_1-1), Holocene transitional loess (L_t), the Middle Holocene palaeosol (S₀), the recent loess (L₀) and the modern topsoil (MS). The Middle Holocene palaeosol (S₀) is separated by three layers of red clay, which were deposited by large flash floods and mudflows during 3960-3650 a BP. These catastrophic events occurred at about 3850 a BP, 3800 a BP and 3600 a BP, respectively. The first mudflow event overflowed the gully channel and spread over a large area on the Yellow River terrace, and covered the eastern part of the large settlement of the Qijia Culture at 3850 a BP. All these facts indicate that the Lajia Ruins was destroyed by earthquakes in combination with large flash floods and mudflows coming along the gullies that originated from the hills of the Tertiary Red Formation. These results are very important in understanding environmental change and human-land relationship evolution, formation of the settlement of the Qijia Culture, pre-historical human impact and causation of catastrophe in the Guanting Basin.

The aeolian deposits in the Yarlung Zangbo catchment are widely developed, and its sedimentary development model is complex. The climatic implication and chronological boundary of the aeolian sediments are not clear enough. Therefore, from the perspective of the whole basin, this paper comprehensively analyzed the development mechanism and mode of the aeolian sediments in different regions. At the same time, the chronological data of aeolian deposition is normalized and compared with the classical curve of global climate. The results show that: (1) The aeolian deposits in the basin have prevailing wind field development patterns and local wind field development patterns, and different patterns of aeolian sediments have different indications for the environment. (2) The aeolian sediments in the basin are mostly deposited in the Last Glacial Maximum (LGM), which mainly developed in different periods since the Late Glacial Period (15 ka BP), and the rich source of alluvial debris may be the main controlling factor of the aeolian activities. The aeolian sediments forming in the earlier time were brought into the river and failed to be preserved due to the warming climate and the erosion transport caused by glacial ablation after the LGM. (3) Since the LGM, the aeolian deposition process of Yarlung Zangbo has been controlled by the regional and global paleoclimate environment, and the fluctuations are drastic. The aeolian sedimentary process of different climate stage have different records to the 30°N summer insolation, the Indian monsoon and westerly winds change signal, but have a better response to the Younger Dryas (YD), global cold event. In short, the sedimentary process of Yarlung Zangbo is similar with that of the other regions on the Tibetan Plateau, and both recorded the global paleoclimatic change, but have its regional uniqueness. 4) The strong accumulation process of aeolian sediments in the Yarlung Zangbo catchment is not a simple correspondence with the global classical climate curve, and it presents a more complex response pattern to global climate change, which shows that the aeolian deposition process of the alpine valley environment is not only affected by the climate of the large area, but also controlled by many local environmental factors. Their relation with the paleoclimatic environment is complex.

Social-Ecological System (SES) is composed of social subsystem, ecological subsystem and the interaction between them; its structure, functions, and characteristics are different from those of social subsystem or ecological subsystem alone. Relying on adaptive social mechanisms of power-sharing and decision making, the adaptive governance of SES aims to guarantee human well-being in a sustainable manner under dynamic conditions. Adaptive governance theory is influenced by "common pool resources management", "resilience" and "governance", and lays a foundation for the construction of transformative governance and collaborative governance. This theory has three main objectives: (1) understanding and coping with the multi-stability, nonlinearity, uncertainty, integrity and complexity of SES; (2) establishing non-confrontational social structure, power-sharing structure and decision-making structure, and match with social subsystem and ecological subsystem; (3) achieving sustainable management of ecosystem services through an integrated approach. Therefore, in the face of the "Anthropocene" of human-behaviour-dominant surface processes, achieving adaptive governance helps to deal with the complexity and uncertainty of SES. Given the rapid changes in China's environment and the increasingly complicated interactions between China and countries all over the world, it will be helpful for future studies to pay close attention to the following fields: (1) understand the multi-interaction processes of a coupled system, and enhance its adaptability; (2) emphasize the significance of a holistic approach of studying SES; (3) improve the ability to understand and predict system dynamics in the context of environmental change.

Previous studies indicate that the lower Yellow River experienced a continuous siltation period and a continuous scouring period during the past 30 years. However, the patterns of main-channel adjustments during these two periods are not clear. Based on measured discharge, sediment load, and cross-sectional data between 1986 and 2015, the changes in the morphological parameters (width, depth, and cross-sectional geomorphic coefficient) of the main channel were analyzed. The results showed that before the operation of the Xiaolangdi Reservoir (XLDR) from 1986-1999, the main channel shrunk continuously with decreasing width and depth. Because the decrease rate in width was greater than that in depth, the geomorphic coef?cient decreased in the reach above Gaocun. In contrast, for the reach below Gaocun, the decrease rate in width was smaller than that in depth, and the geomorphic coef?cient increased. After the XLDR began operating, the main channel eroded continuously, and the width and depth increased from 2000-2015. Because the increase rate in depth was obviously larger than that in width, the geomorphic coef?cient decreased in all sub-reaches. The cross-sectional geometry of the main channel exhibited different adjustment patterns during two periods. Before the XLDR operation, the main channel mainly narrowed in the transverse direction and silted in the vertical direction in the reach above Aishan (AS); in the reach below AS, the main channel primarily vertically silted. After the XLDR operation, the main channel adjusted by widening and deepening in the reach above AS; for the reach below AS, the main channel adjusted mainly by deepening. Compared to the decrease rates in main-channel width and depth during the siltation period, the increase rate in width during the scouring period was obviously smaller, while that in depth was larger. After continuous siltation and scouring, the main-channel cross-sectional geometry changed from relatively wide and shallow to relatively narrow and deep. The pattern of main-channel adjustment was closely related to the water and sediment conditions. For the braided reach, the geomorphic coef?cient was negatively correlated with discharge and positively correlated with suspended sediment concentration (SSC) during the siltation period. In contrast, the geomorphic coefficient was positively correlated with discharge and negatively correlated with SSC during the scouring period. For the transitional and wandering reach, the geomorphic coef?cient was negatively correlated with discharge and positively correlated with SSC.

The global warming hiatus during 1998-2012 has aroused a great public interest in past several years. Based on the air temperature at 622 meteorological stations in China, the response of temperature to global warming hiatus was analyzed on national and regional scales. The main results were as follows: (1) The trend magnitude of air temperature in China was -0.221 ℃/10 a during 1998-2012, which was lower than the long-term trend during 1960-1998 by 0.427 ℃/10 a. There was a warming hiatus in China that was more obvious than the global mean. Winter played a dominant role (contribution rate was 74.13%) in the nationwide warming hiatus, and the contribution of summer was the least among the four seasons. (2) The warming hiatus was spatial incoherent in different climate backgrounds in China. Among the three natural zones in China (the monsoon region of eastern China, the arid region of northwestern China and the high frigid region of Tibetan Plateau), there was a significant cooling in the eastern and northwestern China, especially the eastern China with a contribution rate of 53.79%. In the eastern China, the trend magnitudes were 0.896 ℃/10 a in winter and 0.134 ℃/10 a in summer, respectively. In the Tibetan Plateau, the air temperature has increased by 0.204 ℃/10 a without significant warming hiatus. (3) The warming hiatus in China may be associated with the negative phase of PDO as well as the reduction of sunspot numbers and total solar radiation. (4) Although warming hiatus occurred in China during 1998-2012, the air temperature has rapidly increased after 2012 and is likely to be continuously warming in the next few years.

Non-point source pollution is one of the most severe problems impacting water environments. Identifying potential risk areas and risk paths contributing to non-point source pollution is the soution to this problem. This study introduces the minimum cumulative resistance model of landscape ecology, which is based on land use and soil mapping at a scale of 1∶100000 and DEM data with a resolution of 30 m. The model takes high pollution-loaded cultivated land and construction land as the main sources and uses the Topographic Wetness Index and Runoff Curve Numbers, which can describe the underlying resistance surface runoff yield characteristics, to visually identify and analyze the risk areas and risk paths of the Wuhua River Basin. The results show that underlying surface runoff production results in low-yield flow areas that are mainly concentrated in the southwest of the basin, while high-yield flow areas herringbone throughout the study area. The minimum cumulative resistance model can effectively identify the risk areas and risk paths in this basin. The high-risk areas of non-point source pollution are mainly distributed in Jionglong, Tianxin, Longmu, Tiechang, Dengyun, Tongqu, Heshi, Zishi, Qiling, Huacheng, Zhuanshui, Tanxia and Shuizai, which are located along both sides of the river. The spatial distributions of the risk paths of cultivated land and construction land are significantly different. The effects of cultivated land on water quality of the river are greater than those of construction land on it, and the nutrients and sediments from cultivated land are more likely to run into the receiving water via surface runoff. Vegetation buffer zones should be set up on both sides of the river adjacent to cultivated land when we deal with non-point source pollution that originates from cultivated land, and the harnessment of non-point source pollution originating from construction land should be monitored around major source areas. This study provides a novel method for the identification of source areas and risk paths of non-point source pollution and a theoretical basis to formulate future management strategies.

Soil water is the key link between land surface and atmosphere in water-heat exchange and it is the key element of water cycle. It is also the key control factor affecting the process of surface runoff. The Himalayan-Tibetan Plateau (HTP), also known as the "Asian Water Tower", is the source region of many Asian rivers. Meanwhile, HTP has direct impacts on its surrounding climate via hydro-meteorological processes, and on establishment and maintenance of Asian monsoon. This study collected observed soil moisture data from 100 in-situ soil moisture observatory stations and evaluated applicability of the available remote sensing and reanalysis soil moisture datasets such as ECV, ERA-Interim, MERRA, and Noah at different spatial resolutions (0.25°×0.25°, 0.5°×0.5°, 1°×1°) during different time intervals such as non-freezing and freezing periods. Statistical indicators such as R, RMSE and Bias were used to evaluate the performances of these remote sensing and reanalysis soil moisture datasets. The results indicated that: (1) All remote sensing and reanalysis soil moisture datasets except ERA can well estimate soil moisture changes of the Tibetan Plateau and the soil moisture changes are in generally good line with precipitation changes. In the Naqu region, however, the remote sensing and reanalysis soil moisture datasets substantially underestimate observed soil moisture. In space, MERRA and Noah are mostly consistent with the change of vegetation index, and can well estimate spatial distribution of soil moisture changes. (2) Soil moisture changes across most parts of the Tibetan Plateau are greatly influenced by precipitation changes. In addition, soil moisture changes in the western flank of the Tibetan Plateau and Himalayas are the combined results of melting snow/glaciers and precipitation. (3) Except in the Ngari region, soil moisture during non-freezing period is usually higher than that during freezing period. In the Naqu region, all remote sensing and reanalysis soil moisture datasets overestimate soil moisture amount during freezing periods, while they underestimate it during non-freezing periods. Besides, from a spatial scale viewpoint, at medium and large scales, remote sensing and reanalysis soil moisture datasets can better evaluate soil moisture availability compared with at small scale. This study provides a theoretical basis for selection of the right remote sensing and reanalysis soil moisture datasets for evaluation and analysis of soil moisture of the Tibetan Plateau.

Rice is the staple food in China and its production is impacted jointly by natural environment and human activities. In this process, terrain condition not only determines the spatial pattern of environmental factors, such as water, heat and radiation, but also affects the agricultural management measures. Although many studies focused on the impact of one or several specific factors on crop production, few studies investigated the direct influence of terrain condition on rice production. Therefore, we selected Hunan Province, one of major rice producing areas in China with complex terrain conditions, as the study area. Based on the remote sensing data and statistical data, we applied the spatial statistical analysis to explore the effects of terrain factors on the rice production from the following three aspects: spatial pattern of paddy field, rice production process and the final yield. We found that: (1) Terrain has a significant impact on the spatial distribution of paddy filed at both regional and county scales. Most paddy fields are located on the northern plain and central hills where the elevation is generally below 300 meters with the slope less than 9° and relief degree less than 140 meters. Also, the spatial pattern of paddy fields in Hunan is sensitive to surface roughness and slope position. (2) Terrain does determine the distribution of temperature, sunlight and soil, and these three environmental factors consequently have direct impact on rice growth. Additionally, several terrain factors (elevation, slope and surface roughness) are related with the phenological stage of double-cropping rice, especially for elevation, which is closely associated with the planting stage for early rice and harvesting stage for late rice. (3) However, compared with the pattern of paddy field and rice production process, the influences of terrain factors on the rice yield are not so evident except for elevation. (4) There is a spatial mismatch between spatial distribution of paddy field and production resources due to terrain factors: although paddy fields are widespread in the northern plain, the yield in this region is lower than that in the hilly area of central Hunan due to limited heat. Our results highly imply that the managers should guide farmers to choose suitable variety and planting system and allocate rice production resources in the northern plain so as to ensure food security.

This study, taking the Weihe River Basin in the Guanzhong-Tianshui Economic Region of China as a case, establishes a water conservation ecosystem service network model. Based on Bayesian belief networks, the model forecasts the distribution probability of water conservation ecosystem services projected under different land-use scenarios for the year 2050 with a CA-Marcov model. A key variable subset method is proposed to optimize the spatial pattern of the water conservation ecosystem service. There were three key study findings. First, under the protection scenario, the area of woodland increased by 18.12%, mainly from the conversion of cultivated land. The grassland and cities increased by 0.73% and 0.38%, respectively. The water and unused land were reduced by 5.08% and 0.92%, respectively. The probability of high water conservation value under this scenario is the largest in the three scenarios, and the design of protection scenario is conducive to the formulation of future land use policies. Second, the key factors influencing water conservation ecosystem service include precipitation, evapotranspiration and land use. The state set corresponding to the highest state of water conservation ecosystem service is {precipitation = Highest, evapotranspiration = High, land use = High}, mainly distributed in areas with high annual average rainfall and evapotranspiration and high vegetation coverage. Third, the regions suitable for optimizing water conservation ecosystem service are mainly distributed in the southern part of Maiji District in Tianshui, southwest of Longxian and south of Weibin District in Baoji, northeast of Xunyi County and northwest of Yongshou County in Xianyang, and west of Yaozhou District in Tongchuan. Identifying the optimization regions of water conservation ecosystem service based on Bayesian belief networks, not only helps to develop a better understanding of the water conservation ecosystem services processes, but also increases the rationality of the scenario design and pattern optimization. On this basis, the key variable subset method is crucial to sound eco-environment construction and policy formulation in the study area.