Micrometeorology and water use of mixed crops in upland West Java, Indonesia

The scarcity of information on the micrometeorology and water use of rain-fed humid tropical crops leads to uncertainty about the effects of land cover change on streamflow and atmospheric circulation. We made micrometeorological measurements during five wet season months in a rain-fed cropping system with maize (Zea mays L.) and cassava (Manihot esculenta Crantz) on a bench-terraced hillside in upland West Java, Indonesia. Reflection coefficients varied between 0.07 for moist bare soil and 0.22 for mature cassava with surface mulching. Fluxes of heat and vapour were estimated by the temperature variance method from wet- and dry-bulb temperature fluctuations measured with fast-responding thermocouples. Sensible heat fluxes were 24% larger when determined from Bowen ratios rather than directly, but latent heat fluxes inferred from the two estimates agreed within 10%. The difference was largely attributed to the greater effect of high-frequency losses on directly calculated fluxes. The Penman–Monteith model with optimised values for surface conductance (12.7 mm s−1), roughness length (0.03 times crop height) and displacement length (0.64 times crop height) was used to estimate dry canopy evaporation (Et) for a full year, allowing for the effect of reduced soil water availability during the dry season. Rainfall interception losses (Ei) were estimated with an adapted version of the Gash model calibrated with local measurements of throughfall and stemflow. Simulations with a soil–vegetation–atmosphere transfer model reported elsewhere demonstrated that modelled evaporation rates were consistent with soil hydrological measurements. Total water use was 1228 mm, divided between 1030 mm crop transpiration and soil evaporation and 198 mm rainfall interception losses.