Environmental engineer turning satellite data and water models into decisions for water management.
Water, satellites and data.
I'm an environmental engineer (MSc, Sharif University of Technology) working at the interface of hydrology, satellite remote sensing and machine learning. I build Earth-observation workflows in Google Earth Engine and Python, model surface water and groundwater with WEAP, MODFLOW and MIKE, and turn model results into decision support for water management.
My work on a decision-support system for the Lake Urmia basin, carried out with the University of Delaware, is published in the Journal of Hydrology and was presented at AGU24. I also have hands-on experience in water-quality laboratory analysis and review regularly for Elsevier journals.
Tools and methods.
Research and engineering projects across water resources and Earth observation, each linked to the open-source code that reproduces its methods.
Forty years of Landsat surface-water area for nearly 600 lakes and wetlands. Colour shows each water body's long-term trend; click a point to see its annual water-area curve and fitted trend.
A public Earth Engine app running inside this page. It opens on Lake Razazah (Iraq), which lost about three quarters of its permanent water between 1984–1993 and 2012–2021, with its yearly JRC Global Surface Water animation, area chart and decade comparison. Click any other lake on the map, or search a name, to explore it. The frames take a few seconds to render.
A forecast-informed decision support system for real-time water allocation, coupling a daily WEAP water-balance model with inflow forecasts. Multi-objective evolutionary optimization and AHP were used to compare scenarios; optimized strategies raised environmental water supply by 31% and agricultural profit by 26%.
Multi-year satellite monitoring workflows (Landsat, Sentinel-2) for surface water and land cover in Google Earth Engine and Python, using Random Forest and K-means for classification and change detection, with full accuracy assessment.
Surface-water mapping combining Random Forest classification, K-means clustering and spectral water indices in Google Earth Engine.
MODFLOW simulation of dam leakage and aquifer response to an adjacent river, with analytical and numerical solutions implemented in Python.
Water balance of the sub-basin in ArcGIS and Python, with a chloride (salinity) time series modelled in Python.
Hydrodynamic modelling of oil-slick transport in the Karkheh reservoir using MIKE.
Land-cover classification in ENVI and Python, surface-water mapping with K-means in Google Earth Engine, and watershed delineation in ArcGIS.
Everything below is my own code, public on GitHub under the MIT licence. Each tool has a tutorial page with its demo output, the method, and a step-by-step walkthrough of the source.
9 stepsEarth Engine JS
A public Earth Engine app: click any lake on Earth and watch its JRC Global Surface Water history, with a yearly area chart, a decade comparison and a shareable link. No sign-up.
9 stepsEarth Engine JS
Landsat 5–9 open-water classification on eight spectral water indices, with hold-out accuracy, variable importance and a water-area series since 1984.
7 stepsEarth Engine JS
Monthly NO₂, SO₂, CO, O₃, HCHO (Sentinel-5P) and aerosol optical depth (MODIS) for a region and its stations, in one tidy CSV.
8 stepsEarth Engine JS
Sentinel-2 NDCI chlorophyll-a for lakes and reservoirs with Cloud Score+ masking, monthly lake statistics and a MODIS-Aqua comparison.
7 stepsEarth Engine JS
Monthly precipitation, evaporation, MODIS ET, temperature and GRACE water storage averaged over any river basin, as one table.
5 stepsEarth Engine JS
Forty years of cloud-free summer composites turned into an animated GIF, with the shoreline traced on every frame.
6 lessons · 35 stepsEarth Engine JS
Six commented lessons for students: water frequency, multi-temporal water maps, soil-moisture downscaling and rainfall trend tests.
6 stepsEarth Engine JS
Daily or monthly tropospheric NO₂ from Sentinel-5P TROPOMI at many monitoring stations, with a multi-station chart, a mean NO₂ map and a tidy CSV export.
6 stepsEarth Engine JS
Maps stable cropland, gains and losses at 30 m from the GLAD global cropland maps, reports the area of each class and exports GeoTIFFs.
6 stepsEarth Engine JS
Daily mean, minimum and maximum 2 m temperature from ERA5-Land at any number of stations, exported as one tidy CSV in °C.
6 stepsEarth Engine JS
Samples SRTM or Copernicus DEM every n metres along any line or polyline, charts the terrain profile and exports it as CSV.
Command-line tool for ERA5 monthly-mean humidity, winds and other pressure-level variables from the Climate Data Store, with config files, sub-regions and resumable downloads.
Command-line tool that samples a stack of rasters at points, with automatic reprojection and nodata handling, into one Excel, CSV or GeoPackage table.
Selected outputs.
Meydani, A., Abbasalizadehrezaei, P., Javadian, M., Khodaei, H., Tajrishy, M., Levia, D.F. (2026). Scenario-driven decision support system for assessing water balance to support agricultural and ecosystem sustainability. Journal of Hydrology, 665, 134679.
doi:10.1016/j.jhydrol.2025.134679 →Abbasalizadehrezaei, P., Meydani, A., Khodaei, H., Tajrishy, M., Levia, D.F. (2024). An integrated decision support system for balancing agricultural and environmental water use in the Urmia Lake Basin. AGU24 Annual Meeting, Washington, D.C.
Abbasalizadehrezaei, P. (2026). Twelve open-source tools for Earth-observation, water and climate data (Version 1.0.0) [Software]. Zenodo. MIT licence.
44 reviews for Elsevier: Journal of Environmental Management (35), Agricultural Water Management (8) and Journal of Hydrology (1), covering hydrology, remote sensing and water management.
Research and teaching.
Degrees, distinctions and training.
Thesis: satellite-based assessment of surface water using Random Forest classification, K-means clustering and spectral indices in Google Earth Engine.
Open to research, data and engineering roles and collaborations across Europe.