Differential Impacts of Soil Physical Properties on Root and Leaf Phenology in Cropland

Root phenology is essential for ecosystem cycling and plant resource acquisition, yet its response to soil conditions is less understood than that of leaf phenology. In view of this, the present study examines the effect of soil conditions on both root and leaf phenology in a cropland area over Selhausen in Germany. We developed a hybrid deep-learning model that integrates U-Net and a regression network in this study, which effectively predicted root phenology (R2=0.80–0.92; MAE=0.17–0.90). By leveraging a physically-constrained machine learning approach, we derived leaf phenology from Sentinel satellite imagery and compared leaf phenology with the root initiation and root cessation through root imagery technology (minirhizotrons) across the study region from 2015 to 2021. Results showed that root growth initiation occurred approximately 3 days earlier than leaf phenology, while root cessation was delayed by 25 days under dry soil conditions. Further, we found root phenology to be significantly influenced by soil moisture (β = 0.53, p < 0.001), exhibiting greater sensitivity than soil temperature (β = 0.13, p < 0.001). These findings emphasize the need to distinguish between root and leaf phenological responses in biogeochemical models, highlighting soil heterogeneity as a key factor in under-ground phenological processes.