Responses of surface conductance to forest environments in the Far East

The dependence of surface conductance Gs on various variables in a forest environment is important for characterising the spatiotemporal variations of water, energy and CO2 exchange between the forest and the atmosphere. Using a Jarvis-type model, we examined the variation of Gs in five different mature forests from three climate zones (boreal, cool- and warm-temperate) in the Far East. First, we applied the model using summertime Gs data from each site separately (within-site analysis). We evaluated the maximum surface conductance Gsmax and parameters related to the response of Gs to the environments. We found that these values differed among the locations. Second, we applied the model for pooled Gs data from all the sites and calculated a common parameter set (pooled analysis). In the pooled model, the difference in the observed Gsmax among the sites was expressed as a function of soil water content and the leaf area index. In addition, the responses of Gs to radiation, vapour pressure deficit and air temperature in different forests were also closely represented by common response functions. As a result, the pooled model could estimate the variation of Gs at each site using one parameter set with similar precision to the within-site models. These results suggest that the surface conductance of the various mature forests had the same maximum value and response properties, although we were not able to verify this. Our new parameterisation concept for pooled Gs data should be effective for simultaneous evaluations of the water, energy and CO2 exchanges of forests over wide regions.