On the sensitivity of the Amazon surface climate to two land-surface hydrology schemes using a high-resolution regional climate model (RegCM4)

Two 11-year simulations were conducted to investigate the influence of two runoff schemes in the community land model version 4.5 (CLM45) on the Amazon surface energy balance and surface climate using a high-resolution regional climate model (RegCM4-CLM45). The default scheme is TOPMODEL (TOP), while the alternative is Variable Infiltration Capacity (VIC). In the two simulations, the vegetation status is prescribed (satellite phenology; SP). The first simulation was designated as SP-TOP, while the second simulation was referred to as SP-VIC, and both of them were evaluated using reanalysis products (e.g., ERA5) and micrometeorology data measurements. Results show that the SP-VIC severely underestimates latent heat and overestimates sensible heat fluxes, more than SP-TOP in comparison with the ERA5. This explains the large warm bias observed in the winter season. On the other hand, the SP-VIC shows a slightly smaller dry bias than SP-TOP against the Climate Research Unit (CRU) data. Our results show that SP-VIC does not improve the quality of the simulation compared to SP-TOP, which suggests the necessity of additional calibration of the VIC surface parameters using in situ observations of the Amazon and revising the VIC runoff scheme to perform new sensitivity experiments. The same needs to be done with SP-TOP.