The discipline of land change science has been evolving rapidly in the past decades. Remote sensing played a major role in one of the essential components of land change science, which includes observation, monitoring, and characterization of land change. In this paper, we proposed a new framework of the multifaceted view of land change through the lens of remote sensing and recommended five facets of land change including change location, time, target, metric, and agent. We also evaluated the impacts of spatial, spectral, temporal, angular, and data-integration domains of the remotely sensed data on observing, monitoring, and characterization of different facets of land change, as well as discussed some of the current land change products. We recommend clarifying the specific land change facet being studied in remote sensing of land change, reporting multiple or all facets of land change in remote sensing products, shifting the focus from land cover change to specific change metric and agent, integrating social science data and multi-sensor datasets for a deeper and fuller understanding of land change, and recognizing limitations and weaknesses of remote sensing in land change studies.

We developed a new Object-based Disturbance Agent Classification Approach (ODACA) to characterize land disturbance agents based on Landsat time series. Seven major disturbance agents were characterized, including harvest, mechanical, stress, debris, hydrology, and fire. We first created the land disturbance map by using a modified COntinuous monitoring of Land Disturbance (COLD) algorithm (Zhu et al., 2020), and then established a semi-automated disturbance agent training dataset extraction framework based on existing open-source datasets, with very limited human intervention. The modified COLD algorithm was implemented based on Landsat time series from a single Landsat path to reduce the bidirectional reflectance distribution function effect and issues caused by data density disparity, and the model updating frequency was reduced from every new observation to every three percent of the number of observations used in the previous model updating to improve computational efficiency. Finally, disturbance agents were classified based on ODACA using a Random Forest model with a total of 175 predictor variables that contain rich information in the spectral, temporal, and spatial domains. Accurate land disturbance agent maps were created for five sites in the United States, with an overall accuracy of approximately 99%, and producer’s and user’s accuracies range from 57 to 100%, depending on specific disturbance agents.