Answers are grounded in this paper and cite their sources.
Forest age is an important indicator for forest quality evaluation and a crucial parameter in forest surveys. However, existing studies on forest age estimation are often based on limited field plots and single-date remote sensing data, without fully exploiting long-term repeated survey data, and rarely considering the spatiotemporal correlation of data in the modeling process. To address this gap, we developed a method to estimate the forest age of Pinus kesiya using long-term Landsat imagery and repeated field survey data. Specifically, we first applied the Boruta algorithm to identify important predictors from open-access Landsat and environmental datasets. These variables were then refined to account for spatiotemporal correlations. Subsequently, the INLA-SPDE and Random Forest (RF) models were employed to estimate the forest age of Pinus kesiya forests. The results show that (1) texture information in remote sensing images is an important variable in forest age estimation, and considering the spatiotemporal correlation of remote sensing data can further obtain variables that have significant benefits for forest age estimation; (2) The most accurate forest age fitting results were obtained when a sample dataset from six surveys over a 25-year period was used. As the number of sample survey years decreased, the model accuracy decreased too; (3) When using the INLA-SPDE algorithm to estimate forest age based on repeated survey data, it is necessary to consider the spatiotemporal correlation of the data because it can reduce the estimation error by at least 3 years. The best fitting result was obtained by combining the INLA-SPDE and RF algorithms, with an R² of 0.83 and an RMSE of 6.55 a. This study provides an effective methodological approach for remote sensing estimation of forest age based on long-term repeated survey data, which can provide reliable data support for forest resource monitoring and forest carbon sink management in different regions.
No takes yet. Share an insight, caveat, or question.
Tang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: