Starting Salary: Commensurate with experience
Closing Date: Sunday, February 28, 2021
The NASA Famine Early Warning Systems Network (FEWS NET; https://lis.gsfc.nasa.gov/projects/fewsnet) team seeks a researcher to contribute to the development of an extended outlook crop monitor (EOCM). The EOCM is expected to provide multi-seasonal forecasts (18- to 24-month lead time) of crop conditions and yields. The position involves exploring the optimal use of projections of climate signals, soil moisture, terrestrial water storage (TWS), precipitation, temperature, and other meteorological variables to provide long-lead, multi-seasonal forecasts of a range of terrestrial variables relevant to crop condition and yield assessment. These variables include but are not limited to leaf area index and normalized difference vegetation index (NDVI).
The EOCM postdoctoral researcher position is expected to evaluate statistical, physical, and machine learning modeling approaches to providing long lead time forecasts of terrestrial variables relevant to crop condition and yield assessment. This work will involve evaluating the correlations between forecasted hydrological states (e.g., soil moisture, TWS), climate signals, and meteorological variables with crop condition and yield.
The candidate should have experience applying statistical, physical, and machine learning modeling methods to earth observations. Experience in remote sensing and land surface/hydrologic modeling is highly desired. The candidate should have a PhD in atmospheric science, hydrology, agroclimatology, or related discipline. Proficiency with Python language, Unix, and multiprocessor environments is required. The individual will be expected to present results in peer-reviewed publications and scientific presentations. Experience in organizing meetings, project coordination, and writing and reviewing technical reports is also expected.
To Apply: Interested candidates should send a CV with a list of at least 3 professional references and a cover letter explaining how your qualifications meet the posted requirements to firstname.lastname@example.org. The original posting can be found here.
Starting Salary: Commensurate with experience
Closing Date: Tuesday, February 16, 2021
The successful candidate will work on research, software development and streamlining work-flows related to machine learning applications for crop production forecasting for smallholder agricultural systems at the field to national scales with focus on Sub-saharan Africa. Example projects will involve developing machine learning models to map cropland and crop types from images, identify crop field boundaries, forecast crop yields, and alert of impending production shortfalls. These methods will be used to inform key agricultural and food security decisions as well as develop training materials for a range of public and private stakeholders. This research will be carried out through the use of a wide range of satellite data, unique ground-collected datasets, global archives of diverse socio-economic data, and statistics.
- Must be a U.S. Citizen or Permanent Resident.
- BS or MS in computer science, remote sensing, GIS, geographical sciences, agricultural sciences, physics, engineering, mathematics, or related fields
- Strong Python programming background
- Knowledge of software design principles and tools (e.g., Github/version control)
- Experience using machine learning libraries such as TensorFlow, Keras, PyTorch, and scikit-learn
- Experience developing and/or applying machine learning models for classification, semantic segmentation, and regression
- Experience using remote sensing data, GIS, and applying geospatial algorithms
- Interest in agriculture, food security, and environmental research and applications
- Strong problem-solving and interpersonal skills; ability to work in a small development team environment
- Ability to effectively communicate technical concepts to technical and non-technical staff
- Ability to meet deadlines
- Experience with working Google Earth Engine
- Experience with AWS, Azure, or other cloud computing platforms
- Experience working with medium to high-resolution remote sensing datasets such as Sentinel-2, Sentinel-1, or PlanetScope
- Experience working with geospatial python libraries such as gdal, rasterio, xarray, and geopandas
- Publication and/or scientific writing experience
- Experience with open source GIS software (e.g., QGIS)
- Experience with machine learning methods for small datasets (e.g., semi-supervised methods, few shot learning, weakly supervised methods, etc.)
View additional details and the original posting here.