Significant bias in pond surface temperature measurements caused by solar radiation

Abstract Water temperature is a key abiotic factor routinely measured in ponds and lakes. Here we investigate potential inaccuracies in water temperature measurements caused by the heating of loggers by direct solar radiation. We conducted a comparative analysis between surface temperature measurements obtained from deploying three different logger types in a small pond and a large lake. For temperature loggers with sensing element inside the housing, our results show temperature differences up to 0.8°C between shielded and non-shielded loggers in the small pond. This temperature excess of the non-shielded logger correlates with solar radiation measurements, suggesting that indeed direct solar radiation is the main source of uncertainty for non-shielded loggers at the water surface. In contrast, a more accurate logger with sensing element outside the logger housing was found to remain comparably insensitive to direct solar radiation with temperature difference never exceeding 0.3°C even in the small pond. Finally, measurement differences were always below 0.25°C for all types of loggers in the near-shore of a large lake. We show that using non-shielded temperature loggers to calculate surface heat fluxes may overestimate total heat flux by up to a few tens of Wm⁻ ² on sunny days. Averaged over longer periods, the overestimation caused by using non-shielded loggers can lead to a bias of a few Wm⁻ ². In conclusion, we suggest that cheap but widely used temperature loggers with internal sensing element should be protected from direct sunlight when used in still water bodies.