We developed a linked land-sea modeling framework based on remote sensing and empirical data, which couples sediment export and coral reef models at fine spatial resolution. This spatially-explicit (60 × 60 m) framework simultaneously tracks changes in multiple benthic and fish indicators as a function of land-use and climate change scenarios. We applied this framework in Kubulau District, Fiji, to investigate the effects of logging, agriculture expansion, and restoration on coral reef resilience. Under the deforestation scenario, models projected a 4.5-fold sediment increase (>7,000 t. yr-1) coupled with a significant decrease in benthic habitat quality across 1,940 ha and a reef fish biomass loss of 60.6 t. Under the restoration scenario, models projected a small (<30 t. yr-1) decrease in exported sediments, resulting in a significant increase in benthic habitat quality across 577 ha and a fish biomass gain of 5.7 t. The decrease in benthic habitat quality and loss of fish biomass were greater when combining climate change and deforestation scenarios. We evaluated where land-use change and bleaching scenarios would impact sediment runoff and downstream coral reefs to identify priority areas on land, where conservation or restoration could promote coral reef resilience in the face of climate change.