Extending periodic eddy covariance latent heat fluxes through tree sap-flow measurements to estimate long-term total evaporation in a peat swamp forest

Abstract. A combination of measurement and modelling was used to find a pragmatic solution to estimate the annual total evaporation from the rare and indigenous Nkazana Peat Swamp Forest (PSF) on the east coast of Southern Africa to improve the water balance estimates within the area. Actual total evaporation (ET_a) was measured during three window periods (between 7 and 9 days each) using an eddy covariance (EC) system on a telescopic mast above the forest canopy. Sap flows of an understory tree and an emergent tree were measured using a low-maintenance heat pulse velocity system for an entire hydrological year (October 2009 to September 2010). An empirical model was derived, describing the relationship between ET_a from the Nkazana PSF and sap-flow measurements. These overlapped during two of the window periods (R² = 0.92 and 0.90), providing hourly estimates of ET_a from the Nkazana PSF for a year, totalling 1125 mm (while rainfall was 650 mm). In building the empirical model, it was found that to include the understory tree sap flow provided no benefit to the model performance. In addition, the relationship between the emergent tree sap flow with ET_a between the two field campaigns was consistent and could be represented by a single empirical model (R² = 0.90; RMSE = 0.08 mm h⁻¹).

During the window periods of EC measurement, no single meteorological variable was found to describe the Nkazana PSF ET_a satisfactorily. However, in terms of evaporation models, the hourly FAO Penman–Monteith reference evaporation (ET_o) best described ET_a during the August 2009 (R² = 0.75), November 2009 (R² = 0.85) and March 2010 (R² = 0.76) field campaigns, compared to the Priestley–Taylor potential evaporation (ET_p) model (R² = 0.54, 0.74 and 0.62 during the respective field campaigns). From the extended record of ET_a (derived in this study from sap flow) and ET_o, a monthly crop factor (K_c) was derived for the Nkazana PSF, providing a method of estimating long-term swamp forest water-use from meteorological data. The monthly K_c indicated two distinct periods. From February to May, it was between 1.2 and 1.4 compared with June to January, when the crop factor was 0.8 to 1.0. The derived monthly K_c values were verified as accurate (to one significant digit) using historical data measured at the same site, also using EC, from a previous study.

The measurements provided insights into the microclimate within a subtropical peat swamp forest and the contrasting sap flow of emergent and understory trees. They showed that expensive, high-maintenance equipment can be used during manageable window periods in conjunction with low-maintenance systems, dedicated to individual trees, to derive a model to estimate long-term ET_a over remote heterogeneous forests. In addition, the contrast in annual ET_a and rainfall emphasised the reliance of the Nkazana PSF on groundwater.