Research On The Influence Of Preheating Strategies For Regional Simulation Of Hydrothermal Processes In Permafrost On The Qinghai-Tibet Plateau

Permafrost is the product of long-term interactions of the atmosphere,lithosphere,hydrosphere and others.The periodic freezing and thawing processes of its shallow soil affect the regional hydrological cycle,ecological environment,and infrastructure stability.Permafrost covering nearly a quarter of the land surface in the northern hemisphere is an important part of the land system and a sensitive factor in response to global climate change.Site-scale observations show the global permafrost is undergoing degradation to varing degrees over last decades.The situation of permafrost degradation and its effects attracted large amount of attention from the science community.The Land Surface Models(LSMs)considering the processes of soil freezing and thawing is an important tool to understand the response mechanism of permafrost to climate change.The impacts of initial fields may last for a long time under the effects of soil memories to water and temperature.However,the current permafrost modeling community have limited knowledge of the impacts of spin-up strategies on regional permafrost modeling,and choose strategies based on experience.It leads large uncertainties on simulation results,and promote the urgent need for special study.For the permaforst freezing-thawing processes on the QTP,this study employed an augmented Noah LSM.We investigate the impacts of spin-up strategies consisting of the looping lengths and repeated years on simulation result by utilizing HighPerformance Computing(HPC).In the first group of experiments,the first-year forcing was repeated to construct the forcing data during spin-up period,and the stair-stepping looping lengths were set from 30 to 2,000 years.In the second group,experiments with different repeated years strategies were conducted under sufficient looping lengths,including single-year and multi-year of which the lengths ranging from 5 to 30 years.According to the soil temperature initial field and permafrost area,this study evaluated the influence of spin-up strategies,and its implications on permafrost area change simulation,the length of years to be repeated and scale-dependence were discussed extensively.Finally,we recommended the spin-up strategies on regional permafrost modeling.We conclude that:(1)The underlain permafrost lengthens the model convergence time.The model state at the table of the permafrost is the most difficult to reach convergence.The impacts of the looping length on the initial fields are small.However,in the transition zone between the permafrost and seasonally frozen ground,the tiny differences are amplified in the subsequent simulations and affect the permafrost area dynamics.Different repeated years strategies affect largely the initial fields of soil temperatures,and the influences last for the whole simulation period.(2)The underlain permafrost lengthens the model convergence time.The model state at the table of the permafrost is the most difficult to reach convergence.The impacts of the looping length on the initial fields are small.However,in the transition zone between the permafrost and seasonally frozen ground,the tiny differences are amplified in the subsequent simulations and affect the permafrost area dynamics.Different repeated years strategies affect largely the initial fields of soil temperatures,and the influences last for the whole simulation period.(3)For the simulation of permafrost dynamics by the improved Noah LSM,looping for 500 years is a proper trade-off between computational resources and model convergence on most grids.The single-year repeated strategy may face the problem of representation in the meteorological conditions.The multi-year repeated strategy is an alternative but we should avoid the lengthy repeated years which introduce the trends of climate change.We recommend looping for at least 500 years with the first 5-10 years forcing repeated.This study provides a useful reference for regional permafrost modeling from a new perspective,that is spin-up strategies.Modelers should realize the great uncertainty caused by the spin-up strategies when assessing the response of permafrost to climate change.