Natural and planted forests contribute substantially to regional gross primary productivity (GPP) in subtropical China, yet their long-term productivity trajectories and drought-related GPP responses remain uncertain, particularly across different satellite products. Using two satellite GPP products (MUSES and eLUE), meteorological records, and the Standardized Precipitation Evapotranspiration Index (SPEI), we compared long-term productivity trends in stable natural (796,468 km2) and planted forests (341,696 km2) and evaluated their responses during and after the 2016 summer drought relative to a 2001–2015 baseline. Both forest types experienced significant warming alongside sustained GPP increases; however, trend magnitudes differed across products, with eLUE indicating faster growth in planted forests and MUSES showing identical rates. Despite severe meteorological drought in summer 2016, both products indicated predominantly positive GPP anomalies, with SPEI explaining less than 1.5% of spatial anomaly variation. In the post-drought period (September–October), mean GPP anomalies declined across products, but the resulting status was highly product dependent: MUSES showed stronger negative anomaly changes in planted forests and below-baseline status, whereas eLUE indicated comparable anomaly changes between forest types and persistent above-baseline GPP. In conclusion, these findings highlight product-dependent interpretations of drought-related GPP responses, without establishing universal differences in drought resistance between forest types. These event-specific findings emphasize the need to distinguish meteorological drought exposure, short-term GPP anomaly changes, and productivity relative to historical seasonal baselines, without equating these indicators with long-term ecosystem resilience or recovery.
Sustainable tourism planning in large river basins requires evidence on whether tourism-resource provision, observed visitation, and cross-regional movement coincide. From a recorded-resource–reported-flow–visitor-interest perspective, this study links supply–interest matching with tourist-flow networks in the Upper Yangtze River Basin using Amap point-of-interest datasets and Ctrip and Mafengwo travelogues grouped into three publication-year cohorts (2023–2025). Platform-recorded tourism-resource supply and platform-observed visitor interest were estimated using kernel density estimation with a 30 km search radius and aligned to a 5 km × 5 km grid. Nine matching types were identified using side-specific thresholds fitted to the pooled 2023–2025 distributions and held fixed across cohorts, and reported consecutive visits were reconstructed as directed weighted networks. Recreation represented 43% of recorded supply but 72.39–75.54% of node visits, indicating stronger functional concentration in platform-observed visitor interest. Low-low cells dominated the grid (73.67–75.81%); medium-medium and high-high cells together occupied 5.22–6.14%. Distinguishing these classes shows that the 79.77–81.02% categorical balance primarily reflects low-intensity correspondence. Recorded resource locations formed a core centred on Chengdu–Chongqing and the Sichuan Basin, whereas visitor interest was multicentric and the strongest network nodes clustered in northwestern Yunnan. Within-province transitions dominated; interprovincial transitions represented 11.54–16.56% of total transition weight, with Sichuan–Yunnan and Sichuan–Chongqing forming the main cross-boundary backbone. The combined analysis identifies where recorded resources, visitor interest and itinerary connections converge, informing the selection of areas for visitor monitoring and resource assessment.