Household livelihood strategies are embedded in the natural and socioeconomic contexts in which people live. Analyzing the factors that influence household livelihood choice and defining their consequences can be beneficial for informing rural household policies. In turn, this has great significance for fostering sustainable livelihood strategies. We grouped household livelihood strategies based on the income distribution of 756 households and analyzed their influencing factors and possible livelihood consequences in the watershed of Miyun Reservoir, the only source of surface water currently available for domestic use in Beijing, China. Local farmers’ livelihood strategies can be grouped into three types: farming, local off-farm, and labor-migrant. Farming households have the lowest livelihood capitals, other than natural capital, compared with labor-migrant households and off-farm households, the latter having better livelihood capital status. Geographical location, natural capital, household structure, labor quality, and ecological policies are the main factors affecting farmers’ choice of livelihood strategy. Local off-farm households have a significantly lower dependency on firewood, land resources, and investment than that of farming and labor-migrant households, and have the highest reliance on fossil fuel. This household classification can help us understand the livelihood characteristics, impact factors, and consequences of different types of household strategies, which also suggest tailored policy and management options to promote sustainable livelihoods based on different household types.

Recently, increasingly sophisticated studies have investigated the relationship between agrarian livelihoods and the environment, as well as rural–urban interactions in developing countries. The policies developed to respond to these dynamics can constrain livelihood options or provide additional opportunities. In the present study, using a modified version of the telecoupled sustainable livelihood framework to generalize dynamic livelihood strategies in the context of rural–urban transformation and by focusing on recent research in China, we review important factors that shape rural livelihood strategies as well as the types of strategies that typically intersect with livelihood and environmental dynamics. We then examine telecoupled rural–urban linkages given that the dynamics of the livelihood strategies of farmers can cause flows of labor, capital, ecosystem services, and other processes between rural and urban areas, thereby placing livelihood strategies in a dynamic context, which has not been considered widely in previous research. We show that most previous studies focused on the reduction of environmental impacts via livelihood diversification and rural–urban migration. We propose several areas for future policy development and research.

Household welfare is inseparable from the environment because of its dependence on ecosystems and their services. However, linking household welfare and ecosystems in order to inform differentiated household livelihood development in spatially heterogeneous regions is still a great challenge. Based on a field survey of 1754 households, we grouped the relationships between household welfare (defined by total income) and ecosystem reliance (expressed by an index of dependence on ecosystem services) in the Watershed of Miyun Reservoir, the only source of surface water for domestic use in Beijing, China. The relationships can be grouped into four types: high welfare and high dependency; low welfare and low dependency; high welfare and low dependence; and low welfare and high dependency. Family structure, households’ education and skill level, and the proximity to Beijing have significant impacts on household welfare, while the quantity of natural capital and eco-compensation fund significantly contribute to the ecosystem dependence. Maintaining suitable family size and age structure, improving education and skill levels, and strengthening payment for ecosystem services within low welfare households would be effective approaches to their welfare improvement. The above classification can help design tailored policy and management options to promote sustainable livelihoods based on different household subgroups.

Reliable global solar radiation (R_s) information is crucial for the design and management of solar energy systems for agricultural and industrial production. However, R_s measurements are unavailable in many regions of the world, which impedes the development and application of solar energy. To accurately estimate R_s, this study developed a novel machine learning model, called a Gaussian exponential model (GEM), for daily global R_s estimation. The GEM was compared with four other machine learning models and two empirical models to assess its applicability using daily meteorological data from 1997–2016 from four stations in Northeast China. The results showed that the GEM with complete inputs had the best performance. Machine learning models provided better estimates than empirical models when trained by the same input data. Sunshine duration was the most effective factor determining the accuracy of the machine learning models. Overall, the GEM with complete inputs had the highest accuracy and is recommended for modeling daily R_s in Northeast China.