Simulation Of The Effect Of Water Level Change On Ecosystem CO₂ Exchange Over A Wetland In The Yellow River Delta

As a link between land and sea,the particular hydro-thermal environment of salt marsh wetland makes it one of the major blue carbon ecosystem.Water level is an essential driving factor affecting the blue carbon function of salt marsh wetland ecosystems.Climate change（such as seas level rise and extreme rainfall）have led to variation in water level,which may alter the carbon sink function of salt marsh wetland.However,the response mechanism and simulation of carbon exchange process in salt marsh wetlands to water level change are relatively scarce.In order to evaluate the influence of simulated water level changes on the characteristics of net ecosystem CO₂exchange（NEE）in the salt marsh wetland,this study relies on the field water level control experiment of salt marsh wetland in the Yellow River Delta（natural water level,underground water level of 20 cm,surface water level of 10 cm）.DNDC（denitrification-decomposition）model was used to simulate and predict the effect of water level change on CO₂exchange in salt marsh wetland.We studied the impact of water level changes brought by sea level rise and extreme climate on the carbon exchange process of salt marsh wetland ecosystem.It exploreed the response of atmospheric,vegetation-soil interface CO₂process to water level change.And it provided reference information for the future development of salt marsh wetland carbon storage.It’s helpful to further analyze the response mechanism of salt marsh wetlands to global climate change and human disturbance,and provide reference direction for the protection,management and utilization of salt marsh wetlands.The main conclusions are as follows:During the whole growing season（from May to November）,water level has a significant influence on the vegetation growth and composition in salt marsh wetland,and further affects the release value of R_ecoand the size of NEE.Surface flooding of10 cm stimulated the growth of aquatic plants and resulted in higher biomass,which promoted the release of R_ecoand the absorption of CO₂.However,the 20 cm underground water level inhibited the plant growth,and the corresponding R_ecoand NEE values were both low.During the observation period,the monthly dynamics of R_ecoshow an inverted"V"shape,while that of NEE shows a"U"shape.Within the study range,the maximum net absorption of NEE was 670.63 kg·hm^-2·d^-1,and the maximum net release of R_ecowas 26.86μmol·m^-2·s^-1.The calibrated and validated DNDC model could accurately simulate the effect of water level change on the NEE of the wetland in Yellow River Delta.Daily dynamics of the simulated NEE value are significantly correlated with the field observation results（R²>0.6）.The sensitivity of DNDC model was tested by changing the input parameters such as climate,soil and field management.The response of ecosystem carbon exchange process to changes in average daily temperature,rainfall and water level was the most significant.The effect of water level on NEE was mainly affected by soil respiration（R_s）.It is proved that water level and temperature are the main influencing factors that change the direction and degree of carbon emission,synthesis and decomposition in salt marsh wetlands.Under the climate scenario from 2021 to 2100,the ecosystem carbon exchange process under different water level changes showed different rules with the increase of years.NEE of salt marsh wetland showed a negative value at the natural water level,and the change of water level would significantly affect the NEE of wetland,and then it flattened out with time.It showed that the salt marsh wetland may have certain adaptability to the stable water level change under the future climate conditions.The annual variation of soil organic carbon(d _SOC)and R_ecoshowed an opposite rule in the interannual dynamic variation.d _SOCat the three water levels all showed a fluctuating decreasing trend,indicating that the growth rate of carbon storage in the future salt marsh wetland might decrease with the increase of time.However,the simulation results are not accurate enough to fit the measured values,and the uncertainty of the model still exists widely.