The Improved Freeze–Thaw Process of a Climate‐Vegetation Model: Calibration and Validation Tests in the Source Region of the Yellow River. Issue 23 (13th December 2018)

• Record Type: Journal Article
• Authors: Yang, Q.; Dan, L.; Wu, J.; Jiang, R.; Dan, J.; Li, W.; Yang, F.; Yang, X.; Xia, L.
• Is Part Of: Journal of geophysical research. Volume 123:Issue 23(2018)
• Page Start: 13, 346
• Page End: 13, 367
• Publication Date: 2018-12-13
• Subjects: freeze–thaw process -- the AVIM model -- particle swarm optimization (PSO) method -- calibration -- validation -- source region of the Yellow River; Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1029/2017JD028050
• Languages: English
• ISSNs: 2169-897X

Abstract

Abstract: The freeze–thaw process significantly impacts land surface processes in permafrost and influences the regional climate. In this study, the freeze–thaw process in the atmosphere‐vegetation interaction model (AVIM) was improved by utilizing the universal soil hydrothermal coupling equations, and by introducing the freezing depression point‐based freeze–thaw parameterization scheme to form a model known as the AVIM frozen soil model (AVIM_FSM). Then, the seasonal frozen soil observational data at Maqu station, located in the Yellow River source region, were used to calibrate the freeze–thaw‐related parameters with an artificial intelligence particle swarm optimization method, and the model was validated. The results indicated that by improving the freeze–thaw process, the temporal variations in soil temperature and liquid water content simulated by the AVIM_FSM model agreed well with the observations. By calibrating the parameters, the deviations between the observations and corresponding simulations were reduced compared with those in the AVIM. Based on the AVIM_FSM, the physical mechanism of the freeze–thaw process was discussed, the different freeze–thaw parameterization schemes were compared, the related freeze–thaw process parameters were quantitatively evaluated, and the following was indicated: (1) the freeze–thaw process was mainly determined by radiation, sensible heat flux, and ice change in frozen soil; (2) the freezing depression point‐based freeze–thaw parameterization scheme was superior to the empirical scheme, which can reasonably describe the freezing and thawing start dates with lower deviations; and (3) the particle swarm optimization algorithm can efficiently calibrate the freeze–thaw‐related parameters and improve the simulation accuracy. Key Points: The freeze–thaw process in the vegetation‐climate model AVIM was improved to form a new model known as AVIM_FSM The freeze–thaw‐related parameters were calibrated with the artificial intelligence particle swarm optimization (PSO) method The AVIM_FSM model was validated and compared with the AVIM simulations using seasonal frozen soil observational data