Abstract
Hourly-resolved measurements of stem radial variations (SRVs) provide valuable insights into how climate-induced changes in hydrological regimes affect tree water status and tree stem radial growth. However, while SRVs are easily measured at the individual tree level, currently no methods are available to monitor this phenomenon across broad regions at intra-annual (daily to weekly) scales. Near-surface (in situ) thermal remote sensing-with its sensitivity to plant water status-may provide an approach for monitoring intra-annual SRVs, with the potential for scaling these approaches to the landscape level. Thus, we explored the suitability of in situ thermal remote sensing, in combination with other environmental data, to monitor SRVs in a coniferous forest of the North American Intermountain West. Specifically, we were interested in answering two main questions: Can we use in situ thermal remote sensing by itself and in combination with environmental variables (i.e., photoperiod, photosynthetically active radiation, and soil moisture) to predict (1) daily tree water status and (2) daily tree stem radial growth derived from SRVs? We used data collected by an environmental monitoring network in central Idaho over three growing seasons (2019-2021) to address these questions. Results showed that leaf temperature (T-L) in combination with environmental variables explained up to three-quarters of the SRV-based variability in daily tree water status (in the form of tree water deficit [TWD]) and approximately one-half of the variability in daily stem radial growth. The time of day when T-L was acquired also appeared to change the strength, shape, and predictive power of the models, with acquisition times in the morning and evening showing stronger relationships with daily SRVs than other times of the day. Overall, these results highlight the promise of utilizing thermal remote sensing data to derive tree hydrological and growth status, and reveal key considerations (e.g., the time of data acquisition) for future observational and modeling efforts. This study also provides a benchmark against which to compare future efforts to test these observed relationships at coarser spatial scales.