荒漠草原生长季不同时间尺度水热通量变化特征及其影响因子

荒漠草原水热通量变化特征及其影响因素分析对提高区域水热交换规律认识和了解其在气候变化中的作用具有重要意义。该研究基于内蒙古荒漠草原2018和2019年4—10月涡度相关系统与相应的环境因子观测资料，分析了荒漠草原不同时间尺度水热通量（感热通量（sensible heat flux，H）和潜热通量（latent heat flux，LE））及能量占比（H/LE）的变化特征及其环境因子调控机制。结果表明：1）H为该生态系统的主要能量消耗形式，且同一环境因子在不同生育期对H、LE、H/LE的影响存在差异；2）小时尺度上，H、LE呈单峰型变化，峰值均主要出现在12：00—14：00之间，且LE峰值出现时间较H滞后1～2 h。H的主要影响因素为平均风速（wind speed，Ws），相对湿度（relative humidity，RH）、饱和水汽压差（vapor pressure deficit，VPD）和空气温度（air temperature，Ta），相关系数（correlation coefficient，r）绝对值大小分别为Ws（0.72）> RH（0.66）> VPD（0.61）>Ta（0.38）。LE的主要影响因素为Ta、VPD、地面温度（soil temperature，Ts）和降雨（precipitation，P）；3）日尺度上，H、LE呈明显的季节变化，H的主要影响因素是RH、Ws、Ts、Ta、P和VPD，H/LE主要影响因素为RH、Ws、Ts和Ta；4）月尺度上，H、LE近似呈单峰型变化，夏季能量消散方式以LE为主，H的主要影响因素为RH、Ws和VPD，其中VPD正效应最强（r=0.75），P对LE的正效应最强（r=0.75），而RH对H/LE的负效应最强（r=-0.76）。研究结果有利于准确理解多时间尺度下荒漠草原H和LE对环境因子的响应，同时可为草原生态功能区的植被恢复、生态环境保护提供理论依据和技术支撑。

Characteristics and impact factors of water and heat flux changes at different time scales during the growing season of desert grasslands

Water and heat flux is a key components for the energy exchange of surface air in atmospheric circulation. The driving mechanism can include the sensible heat flux (H) and latent heat flux (LE). There is a great variation in the water and heat flux in desert grasslands. The influencing factors are of great significance to understanding the regional exchange of water and heat flux against climate change. This study aims to explore the variation patterns of water and heat flux at different time scales. The eddy covariance and continuous observation were also utilized from automatic meteorological stations in Xilamuren Town, Damao Banner, Inner Mongolia, China. After that, the regulatory mechanisms of environmental factors were proposed for the desert grassland during growing seasons (April to October) in 2018 and 2019. A systematic analysis was made on the H, LE, and energy ratio (H/LE) in the desert grasslands at different time scales, together with their environmental control mechanisms. The results indicated that the H of plants was the main form of near-surface energy consumption in the desert grassland during the growth season. There was an unimodal variation in the H and LE on an hourly scale. The H and LE peaks mainly occurred between 12:00 and 14:00, where the time of LE peak occurrence was lagged by 1-2 h, compared with the H. The variation range of H and LE between 8:00—18:00 was 0.1-160.5 and 15.4-90.9 W/m², respectively. The main influencing factors of H were the relative humidity (RH), wind speed (Ws), and vapor pressure deficit (VPD), with the correlation coefficients (r) of -0.66, 0.72, and 0.61, respectively. The main influencing factors of LE were the RH, Ws, soil temperature (Ts), and VPD, with correlation coefficients of -0.074, 0.032, 0.61, and 0.67, respectively. The main influencing factors of H/LE were the RH, Ws, Ts, and VPD, with correlation coefficients of -0.74, 0.86, -0.23, and 0.24, respectively. There was an outstanding seasonal variation in the H and LE on a daily scale. The maximum is also reached before and after the middle stage of plant growth. The daily variation ranges of H, LE, and H/LE during the plant growth season were 4.2-186.1, 1.3-160.9 W/m², and 0.18-26.50, respectively. The main influencing factors of H were the RH, Ws, air temperature (Ta), precipitation (P), and VPD, with correlation coefficients of 0.13, 0.21, -0.15, and 0.39, respectively. The main influencing factors of LE were the RH, Ws, and Ta, where there was no significant correlation with the Ts and P. The main influencing factor of H/LE was the RH with a correlation coefficient of -0.16, which was not significantly correlated with other factors. On a monthly scale, the H and LE were approximated a unimodal change. The average H and LE reached the maximum in June and August. The variation ranged of H and LE were 36.4-133.8 and 15.6-96.3 W/m², respectively, from April to October. The LE was the main mode of energy dissipation in summer. The main influencing factors of H were the Ts, Ta, and VPD, where there was the strongest positive correlation of VPD. The LE was significantly correlated with the RH, Ws, Ts, Ta, and P, where the P shared the strongest positive correlation (r =0.75). There was a significant correlation between H/LE with Ts, Ta, P, and VPD, in which the Ts shared the strongest positive correlation (r =0.76). The finding can provide the theoretical basis and practical significance for vegetation restoration and environment protection in ecological functional grassland.