MEASURING SOLAR-RADIATION TRANSMISSION IN TROPICAL AGRICULTURE USING TUBE SOLARIMETERS - A WARNING

In Kenya, the ratio of the output of north-south (N-S) mounted tube so larimeters (TSLs) to that of dome solarimeters and an east-west (E-W) mounted TSL confirmed large divergences of the N-S mounted instrument around solar noon and early and late in the day, owing to direction-re lated geometrical factors of radiation interaction with the glass enve lope, which are not fully understood. A difference in solarimeter outp ut of NE-SW mounting compared with E-W mounting remained within the ac curacy limits. High ambient temperatures (near 40 degrees C) with low wind speeds (less than 0.3 m s(-1)) appeared to increase tube output r elative to that of the Kipp solarimeter by 10-20%. Condensation inside the tube increased the output by less than 5%. Ancillary indoor data in the Netherlands confirm these ambient effects. These errors may add to the geometrical ones and together they lead to instrument-based di stortions in the interpretation of derived parameters such as radiant energy conversion to dry matter and radiation extinction in mulches an d crops. An example of the latter, for grass mulches, is given that co nfirms earlier results obtained in Dar es Salaam. Calibrations of the tubes as a function of time and ambient conditions against a standard instrument and, where possible, near E-W mounting are therefore necess ary under tropical conditions. Under these conditions, filtered TSLs a ppear to be unreliable for determination of photosynthetically active radiation, as a comparison with a Line Quantum Sensor confirmed. (C) 1 997 Elsevier Science B.V.