摘要
Forest plantations for wood production are an increasingly important land use in southern Australia, and there are potentially important hydrologic consequences of what is mostly a change in land use from agriculture to silviculture. An ability to predict, with some degree of accuracy, the impact of plantation expansion on surface water and groundwater resources is essential. A validated process-based modelling approach, integrating the many interacting environmental and management factors which may influence plantation growth and transpiration, can be used for this purpose. The 3PG forest growth model has been evaluated for a number of species from widely differing climate and site conditions. While growth predictions have been validated, little attention has been given to testing the accuracy of the transpiration predictions or the model's representation of the water balance. We enhanced the 3PG forest growth model (known as 3PG ) and then integrated it into the Catchment Analysis Tool (CAT), so that it now interfaces with a more detailed multi-layered, daily time step representation of the soil water balance. Simulated transpiration using 3PG in CAT was compared with field measurements in 30 plots (across 15 sites) representing 5 common plantation species (Eucalyptus globulus, E. nitens, E. grandis, E. regnans and Pinus radiata) across ages 2–31 years. Mean daily plot transpiration during the measurement periods ranged between 0.4 and 4.2 mm day−1 (average 2.0 mm day−1). Simulated mean daily plot transpiration using 3PG in CAT for Eucalyptus was good (coefficient of efficiency = 0.80; R2 = 0.81). While the model tended to slightly under-predict transpiration at higher measured rates (>3.5 mm day−1), predictions at monthly timescales had acceptable accuracy. The integration of 3PG into CAT resulted in an improvement in accuracy and applicability of CAT, and provides for the spatial application of 3PG across diverse and mixed land use catchments for investigation into carbon and water movement in forest systems.