Farmers are among the most vulnerable groups that need to adapt to climate change. Correct perception is a prerequisite for farmers to adopt adaptation strategies, which plays a crucial guiding role in the development of adaptation plans and the improvement of the security of livelihoods. This study aimed to compare farmers’ perceptions of temperature and precipitation change with meteorological data in two regions of the Tibetan Plateau, analyzed how irrigation affects farmers’ perceptions. Data were obtained from local meteorological stations and household questionnaires (N = 1005). The study found that, since 1987, the climate warming trend was significant (p < 0.01), and the temperature increase was faster in winter. Precipitation had no significant change trend, but the seasonal variations indicated that the precipitation concentration period moved forward in the Pumqu River Basin and was delayed a month in the Yellow River-Huangshui River valley. The farmers’ perception of temperature change was consistent with meteorological data, but there was an obvious difference in precipitation perception between the two regions. We noticed that irrigation facilities played a mediating role on precipitation perception and farmers having access to irrigation facilities were more likely to perceive increased precipitation. Finally, this study suggested that meteorological data and farmers’ perceptions should be integrated when developing policies, rather than just considering actual climate trends. Simultaneously, while strengthening irrigation investment, the government should also pay attention to publicizing the consequences of climate change and improving farmers’ abilities of risk perception.

Rapid urbanisation in China has led to massive outmigration in rural regions, which has changed the regional labour force structure and can have various profound impacts as a result. This research used a case study in Southwest China to investigate how regional land use patterns have been changed in the context of rural outmigration and assessed the resulting dynamics on local ecological environment. The key findings include: (1) The local land conversion process was mainly characterised by the conversion of farmland (−18.3%) to residential area (+268.3%) and woodland (+55.6%) during 2000–2018; (2) about 83.7% of area showed a statistically significant increase in the normalised difference vegetation index (NDVI), which was not due to human interference factors (e.g., afforestation). Correlation analyses showed that depopulation (R = −0.514, p < 0.01) and local mild temperature (R = 0.505, p < 0.01) could be the main contributors. Only 2.5% of the area had decreased NDVI and this was directly caused by human activities (e.g., urban area expansion). These results implied that vegetation improvement can occur in the context of depopulation and farmland reduction, which did not significantly threaten the local agricultural sector. It then could be a good choice to allow those high-slope and biophysically poor farmlands to undergo forest succession rather than cultivation. Farmers in Southwest China should make a full use of the existing low-slope arable land to curb the declining trend of local farmland amount, in order to meet the future challenges brought by urbanisation. Enhanced agricultural infrastructure, mechanised farming and guide from local government can help achieve this goal. This study provided new insights and more realistic scenarios for rural development in Southwest China. The research findings are expected to provide a better understanding to enable sustainable land use management in Southwest China.