Measurement and modelling of evaporation from a coastal wetland in Maputaland, South Africa

Abstract. The surface renewal (SR) method was used to determine the long-term (12 months) total evaporation (ET) from the Mfabeni Mire with calibration using eddy covariance during two window periods of approximately one week each. The SR method was found to be inexpensive, reliable and with low power requirements for unattended operation.

Despite maximum ET rates of up to 6.0 mm day⁻¹, the average summer (October to March) ET was lower (3.2 mm day⁻¹) due to early morning cloud cover that persisted until nearly midday at times. This reduced the daily available energy, and the ET was lower than expected despite the available water and high average wind speeds. In winter (May to September), there was less cloud cover but the average ET was only 1.8 mm day⁻¹ due to plant senescence. In general ET was suppressed by the inflow of humid air (low vapour pressure deficit) and the comparatively low leaf area index of the wetland vegetation. The accumulated ET over 12 months was 900 mm. Daily ET estimates were compared to the Priestley-Taylor model results and a calibration α = 1.0 (R² = 0.96) was obtained for the site. A monthly crop factor (K_c) was determined for the standardised FAO-56 Penman-Monteith. However, K_c was variable in some months and should be used with caution for daily ET modelling.

These results represent not only some of the first long-term measurements of ET from a wetland in southern Africa, but also one of the few studies of actual ET in a subtropical peatland in the Southern Hemisphere. The study provides wetland ecologists and hydrologists with guidelines for the use of two internationally applied models for the estimation of wetland ET within a coastal, subtropical environment and shows that wetlands are not necessarily high water users.