Simulation Of Forests Transpiration In China Based On Plant Hydraulics

Vegetation transpiration connects water,carbon,and energy cycles through stomatal behavior.Exploring the dynamic change mechanism of transpiration helps deepen the understanding of vegetation ecosystem processes and provides a scientific theoretical basis for carbon peak and carbon neutrality implementation path.As an important nexus between hydrology and ecology,plant hydraulics studies water movement law in the soil-plant-atmosphere continuum(SPAC)system.It describes the mechanisms of water absorption,transportation,storage,and utilization by plants,revealing the biophysical mechanism of transpiration.However,traditional transpiration simulations are usually based on empirical models without plant hydraulic processes,characterizing plant water stress according to soil hydraulics.The limitation of these models is increasing under climate change,thereby promoting the rapid development of plant hydraulics mechanism modeling in recent years.A few studies have proved that physically-based plant hydraulic models can significantly improve transpiration simulation accuracy under drought conditions compared with empirically-based soil hydraulic models.Under the global warming background,increases in atmospheric evaporative demand and frequency and intensity of soil droughts make vegetation face potentially enhanced water stress,and the regulation and impacts of plant hydraulic processes on transpiration are also likely more significant.For a more comprehensive acknowledgment of the climatic drivers of transpiration changes and to clarify the regulation mechanism of plant hydraulic processes to transpiration and its impacts on other hydrological variables,this dissertation adopts the forests within China defined by the Moderate Resolution Imaging Spectroradiometer(MODIS)land cover type product as the study regions,comprehensively investigating the responses of transpiration to atmospheric and(or)soil water stresses as well as the impacts of plant water-use strategy on key terrestrial hydrological variables based on plant hydraulics.The main results are as follows:(1)A machine learning model is constructed to estimate transpiration based on transpiration product and meteorological factors,and a series of single-factor detrending simulation experiments are carried out to reveal the climatic drivers of transpiration changes over the forest regions of China.At each forest grid in China,a regression model through a random forest algorithm for predicting transpiration by meteorological factors is constructed based on the series from 1981 to 2015 of growing season averages of transpiration product,solar radiation,air temperature,VPD,and precipitation.On the basis of this model’s framework,the effect of a single meteorological factor on transpiration is separated;thereby,the quantitative attribution of trends in transpiration is conducted.Also,the partial correlation analysis between meteorological factors and transpiration is performed to consolidate the conclusion.The result shows a turning point in 2000 in the growing season averaged transpiration series,changing from an insignificant trend during the previous period to a significant downward trend from2000 to 2015,whose dominant factor is VPD shown by attribution analysis.The mechanism of VPD effect on transpiration is different in different climatic regions.In arid and semi-arid regions,the significant rise of VPD is the leading cause of the significant decrease in transpiration since the 21 st century.This is because the transpiration in these regions is mainly limited by soil moisture,and vegetation is susceptible to atmospheric water stress under high VPD,thereby reducing the transpiration water loss by reducing stomatal aperture.While in humid and sub-humid regions,soil water supply to plants is relatively sufficient,the significant decline of VPD there from 2000 to 2015 reduces the driving force of plant water transfer to the atmosphere,thereby leading to a decrease in transpiration.(2)Based on the Noah-MP land surface model,the response of plant hydraulic processes to water stresses is comprehensively recognized by comparing the transpiration and evapotranspiration simulations differences between plant hydraulics schemes and soil hydraulics schemes under a variety of contexts,including different water stresses,climatic regions,and time scales.Under four different water stress conditions defined by the 5th/95 th percentiles of VPD and soil moisture(SM),the differences in transpiration and evapotranspiration between plant hydraulics and soil hydraulics are investigated,taking the Noah,CLM,and SSi B three soil hydraulics schemes as a reference.1)The daily-scale result shows that the response of transpiration to atmospheric and(or)soil water stresses is related to local climate types.For atmospheric water stress,in arid and semi-arid regions,the regulation of vegetation on transpiration mainly depends on the degree of atmospheric water deficit,and transpiration is sensitive to atmospheric water stress;in sub-humid regions,the response of transpiration to atmospheric water stress depends on soil water status,only under low SM condition would transpiration be greatly affected by atmospheric water stress;in humid regions,the relatively sufficient water supply of soil to plants results in a limited constraint of atmospheric water stress on transpiration.For soil water stress,in arid,semi-arid,and sub-humid regions,the transpiration simulations based on plant hydraulics scheme under soil water stress alone are larger than that based on soil hydraulics scheme due to the characterization of plant hydraulic capacity;moreover,in humid regions,stem water storage can provide additional water for leaf transpiration even under the soil and atmospheric combined water stress.2)The diel hourly-scale result shows that the plant hydraulics scheme can reflect night-time plant water recharge and the response of stomata to atmospheric water stress at noon,thereby simulating more realistic diurnal variations of transpiration and evapotranspiration.Compared with the soil hydraulics scheme,the plant hydraulics scheme can not only represent the decline in transpiration at noon in response to high VPD but also represent the supplement of stem water storage to transpiration water consumption when the soil water supply is insufficient.(3)Based on the Noah-MP land surface model configured with the plant hydraulics scheme,the effects of “risk-averse” and “risk-prone” plant hydraulic regulation strategies on the simulations of transpiration and other key hydrological variables under normal climate and drought stress are comprehensively investigated.Respectively taking isohydric maple and anisohydric oak as the typical representative of risk-averse and risk-prone plants,their hydraulic parameters verified at tree level are input in plant hydraulics scheme for simulation experiments and further comparative analysis.The results show that transpiration simulations based on different plant water-use strategies are opposite under different climate conditions.Under normal climate,risk-averse vegetation has larger root water uptake due to its greater root area index in shallow soil,thereby having higher transpiration than risk-prone vegetation.While under drought stress,risk-prone vegetation absorbs more water from the deep soil when the shallow soil is dry using its deeper roots,thereby having higher transpiration than risk-averse vegetation.As the budgets of underground water,the difference in terrestrial water storage changes and underground runoff are affected by the setting of different soil column thicknesses to some extent.In the rainy season,compared with the risk-averse scenario with only 2 m soil layer,more precipitation is stored on land in the risk-prone scenario with 10 m soil layer,leading to a larger positive anomaly of terrestrial water storage and lower underground runoff;the situation is reversed in the dry season.