基于Penman-Monteith方程模拟青海云杉生长季日蒸腾过程

青海云杉（Picea crassifolia）作为我国黄土高原与青藏高原地区的主要造林树种，对其林分蒸腾耗水特征的研究，能够更合理的指导该地区植被重建与林分调控，以加强林分的稳定性，提高水分利用效率。为了揭示青海云杉在生长季内的冠层蒸腾规律以及冠层整体气孔阻力与环境因子的响应，评价Penman-Monteith方程在青海云杉冠层尺度上的适用性，采用探针式热扩散茎流计（TDP）进行测定，同步长期监测了环境数据，利用反推法建立冠层整体气孔阻力（r_sT）与环境因子之间的回归模型，结合Penman-Monteith方程模拟出青海云杉生长季的日蒸腾量，采用均方根误差、平均绝对误差和平均相对误差对蒸腾量的实测值与模拟值进行误差分析，以验证模型的准确性。得出的主要结论有：（1）生长季内青海云杉日蒸腾量随月份变化呈先增高后降低的趋势；各月蒸腾量占潜在蒸散量的比例为7月（79.68%） > 8月（72.71%） > 6月（72.53%） > 5月（67.08%） > 9月（66.48%） > 10月（64.29%）；（2）树干液流对气象因子的滞后时间为0.5 h；（3）不同月份云杉冠层整体气孔阻力（r_sT）与空气相对湿度（RH）呈正相关关系，与大气温度（T）、饱和水汽压差（VPD）呈负相关关系；（4）应用所建立的多因素回归模型结合Penman-Monteith方程对青海云杉蒸腾量进行模拟验证，累计平均相对误差为14.381%，平均绝对误差为0.160 mm，均方根误差为0.2。综上所述，Penman-Monteith方程在林分冠层尺度上有较好适用性根据所建立的多因素回归模型并结合Penman-Monteith方程，可以利用饱和水汽压差、温度和空气相对湿度三个气象因子较好地模拟日蒸腾过程。

Simulation of daily transpiration of Picea crassifolia in growing season based on Penman-Monteith equation

Picea crassifolia is the main afforestation tree species in the Loess Plateau in China. Vegetation reconstruction and forest stand regulation can be guided reasonably by studying water consumption characteristics in this area. It is important for forest stand to strengthen stability and to improve water use efficiency (WUE). In order to reveal the canopy transpiration characteristics of Picea crassifolia during the growing season and the response of the overall canopy stomatal resistance to environmental factors, and to understand the equation of Penman-Monteith whether or not suitable for scale of canopy of Picea crassifolia, we used thermal dissipation probe (TDP) to measure sap flow of Picea crassifolia and monitored the factor of environment by weather station (DAVIS). Then we framed the regression model between Canopy overall stomatal resistance (r_sT) and the environmental factors by inverse method. Lastly, we used equation of Penman-Monteith to predict the transpiration of Picea crassifolia during the growing season. The Root Mean Squared Error, Mean Absolute Deviation and Mean Relative Error would be the indexes to test and verify the accuracy of model we framed when we compared the predicted value with the measured value. The results show that (1) the daily evapotranspiration of Picea crassifolia increases firstly and then decreases with the monthly changes during the growth season. The ratio of monthly evapotranspiration to potential evapotranspiration is Jul (79.68%) > Aug (72.71%)> Jun (72.53%) > May (67.08%) > Sep (66.48%) > Oct (64.29%). (2) There is an obvious Lag-time phenomenon between sap flow and meteorological factors about 0.5 h. (3) The r_sT is positively correlated with relative humidity (RH), and negatively correlated with atmospheric temperature (T) and saturated water vapor pressure difference (VPD). (4) We applied the established multi-factor regression model and Penman-Monteith equation to simulate and verify transpiration of Picea crassifolia. The results indicated that the cumulatively average relative error is 14.381%, the average absolute error is 0.160 mm, and the root mean square error is 0.2. The simulation effect is basically feasible. In summary, the equation of Penman-Monteith has good applicability on the stand canopy scale. We can accord to the established multi-factor regression model combined with the equation of Penman-Monteith, three meteorological factors of saturated water VPD, T and RH can be used to simulate the evapotranspiration on any day, avoiding the inconvenience of long-term observation.