Existing land surface models (LSMs) describe physical and biological processes that occur over a wide range of spatial and temporal scales. In addition, many processes within LSMs are nonlinearly coupled and therefore simple linear upscaling techniques can result in large prediction error. We studied the application of Proper Orthogonal Decomposition mapping method that reconstructs temporally-resolved fine-resolution solutions based on coarse-resolution solutions. For study sites in a polygonal tundra landscape near Barrow, Alaska, the results indicate that the ROM produced a significant computational speedup (>1000) with very small relative approximation error (<0.1%) for two validation years not used in training the ROM. We also demonstrated that our approach: (1) efficiently corrects for coarse-resolution model bias and (2) can be used for polygonal tundra sites not included in the training dataset with relatively good accuracy (< 1.5% relative error), thereby allowing for the possibility of applying these ROMs across a much larger landscape. We also looked at regional and watershed scale models where the solutions are significantly more heterogeneous and preliminary results indicated that that this method has the potential to efficiently increase the resolution of land models for coupled climate simulations, allowing LSMs to be used at spatial scales consistent with mechanistic physical process representation.