Linking forest management to groundwater recharge: modeling forest floor evaporation in plantation forests

In Japan, plantation forests have accumulated substantial timber resources; however, fragmented ownership and declining timber prices have resulted in management shortfalls.
Kumamoto Prefecture has enacted an ordinance requiring permit holders in designated priority areas to implement groundwater recharge equivalent to their extraction volume (in principle, 100%) to ensure sustainable groundwater use. If increases in water resources resulting from forest thinning can be scientifically quantified, this ordinance could provide an incentive for forest management under this framework. Currently, groundwater recharge is estimated using a simple coefficient-based formula, and quantitative evaluation remains limited.

Forest water-balance studies have extensively examined canopy interception and tree transpiration; however, knowledge of forest floor evaporation remains limited, despite its importance for accurately assessing changes following thinning. Lysimeters provide high-accuracy measurements of forest floor evaporation but are difficult to operate unattended over long periods. Soil-moisture sensors allow stable continuous measurements but have limited spatial representativeness. In contrast, the eddy covariance method enables non-destructive, wide-area flux observation. While eddy covariance studies have recently been conducted in forests with low tree density, applications in dense forests remain scarce. This study aims to develop a simplified measurement approach and an estimation model for forest floor evaporation to improve water-balance evaluation in managed forests.

Field experiments were conducted during summer (June 28–September 30) in Japanese cypress plantations with stand densities of 454, 927, and 1268 trees ha⁻¹. Forest floor evaporation was measured using lysimeters and an LI-710 evapotranspiration sensor based on a simplified eddy covariance approach. As a simplified method, soil-moisture sensors were installed at depths of 5, 10, and 15 cm. Hourly evaporation rates were derived from each dataset, evaporation between rainfall events was compared, and an estimation model was examined using the measurements together with forest microclimate data.

Across observation periods, forest floor evaporation showed broadly similar temporal patterns among lysimeter, soil-moisture-derived, and simplified eddy-covariance estimates. This consistency suggests that forest floor evaporation can be reasonably represented using simplified measurements under different forest management conditions. Although seasonal variability was not assessed, the results indicate the applicability of sensor-based estimation methods and their usefulness for improving water-balance evaluation related to groundwater recharge in managed forests.