Postdoctoral fellow in Forest Remote Sensing and Climate Risk Assessment at Mgeo

Descrizione

Testo originale dell'annuncio (in inglese)

Subject description Forests in Sweden face growing climate-related pressures: intensifying droughts, more frequent and damaging storms, and cascading risks such as bark-beetle outbreaks. Forest owners and regional authorities need spatially detailed, timely information on where climate stress is emerging, and which stands are most exposed, but current monitoring relies largely on field inspections and coarse regional assessments that are often too slow or too limited for operational decision-making. The Department of Earth and Environmental Sciences announces a postdoctoral position within a project that addresses this gap. The project aims to develop an operational, web-based service that uses Landsat and Sentinel satellite observations, tree-species maps, canopy structure data, and climate indicators to provide continuous climate-risk assessments for Sweden's forests. The service will deliver drought-stress maps by species, a storm-exposure hazard index, and bark-beetle susceptibility signals, designed in collaboration with regional authorities and forest owners associations. As part of this work, the project will investigate the spatial and temporal mechanisms linking windthrow, drought, and bark beetle outbreaks, providing the scientific foundation for the service. Work duties The main duties involved in a postdoctoral position are to conduct research. Teaching may also be included depending on the progress of the postdoctoral project but must not take up more than 20% of working hours. In connection to this, the position includes the opportunity for three weeks of training in higher education teaching and learning. The postdoctoral fellow will: Develop and maintain harmonized satellite time-series datasets (Landsat and Sentinel-2) and integrate them with ancillary data including airborne lidar, tree-species maps, and gridded climate fields. Produce spatially explicit climate risk layers, including drought-stress maps by species, a storm-exposure hazard index, and bark-beetle susceptibility signals. Contribute to the development of an operational web-based monitoring service, including automated update workflows and a user-facing web GIS dashboard. Investigate the forest, landscape, and climate conditions that drive storm susceptibility, using major windstorms as natural experiments and interpretable machine-/deep-learning models. Publish research results in high-quality international journals (at least two peer-reviewed papers are expected). Eligibility Appointment to a postdoctoral position at Lund University requires that the applicant has a PhD, or an international degree deemed equivalent to a PhD, within the subject of the position, completed no more than three years before the date of the employment decision. Under special circumstances, the doctoral degree can have been completed earlier. Additional requirements: PhD in physical geography, remote sensing, geoinformatics, forest ecology, environmental science, or a closely related field. Demonstrated experience with satellite remote sensing for large-scale vegetation or forest monitoring, including Landsat and/or Sentinel-1/2 time series. Strong proficiency in Python for geospatial data analysis and machine learning (e.g. XGBoost), including model interpretation techniques (e.g. SHAP). Very good oral and written proficiency in English. Excellent communication skills, ability to work both independently and in a team, and to complete deliverables on time. A publication record in peer-reviewed international journals. Assessment criteria This is a career-development position primarily focused on research. The position is intended as an initial step in a research career, and assessment will primarily be based on scientific merits and potential as an independent researcher. Particular emphasis will be placed on scientific ability within the subject area. The following knowledge areas will form the basis for assessment: Quantitative remote sensing, includin