dc.description.abstract | It is helpful to improve long-term climate simulation more reality by using land-surface parameterization. The land-surface observations are very short of in Taiwan, and it almost has no applicable data for simulation using. In this study, we utilize off-line NOAH LSM to simulate land surface and soil data of Taiwan, then put in MM5 with low boundary condition to simulate climate from May to Aug in 2001. Compare the simulation result with ECMWF reanalysis data to find the impact of using these soil data. There are four experiments with different soil data, data assimilation, and initiation soil temperature, to study the interaction between atmosphere and soil.
The off-line NOAH LSM could simulate correct soil temperature annual change, but the temperature values are in general lower than the observations. The soil moisture might be too wet for a long term. Owing to the climate difference between Great Plain and Taiwan, the simulation of off-line NOAH LSM makes bias. It is difficult to amend NOAH LSM because the soil, vegetation, and surface condition observations are lack.
The most bias of temperature field could be found lower bias near the Yellow Sea and higher bias near the tropic ocean area. It is due to cumulus parameterization and more moist simulation value. The lower soil temperature and wetter soil moisture in the simulation might make air and surface temperature lower than observation over Taiwan. It still needs to improve on simulating the strength of typhoons and stationary front events.
Lower soil temperature initiation would make the air and surface temperature lower is due to the interaction between soil and atmosphere. The thermal conduction in soil has time lag in deep soil layer, so the different soil layers have different time intervals for lower temperature values restoring in Taiwan. The lower air temperature over Taiwan would be taken by wind, and makes the temperature field bias display the northeast to southwest wave state. | en_US |