Projects
I've actively participated in writing and collaboration in various multidisciplinary projects including Digital Twin Earth (DTE) Hydrology Evolution, 4Dhydro ,4D MED-Hydrology, extrAIm, and Open-Earth-Monitor Cyberinfrastructure (OEMC).
HR-precipNet
A Machine Learning Framework for 1-km High-Resolution Satellite Precipitation Estimation
I developed a A Machine Learning Framework to estimate highest spatial resolution (1km) Satellite Precipitation products. In this framwrok we evaluate several U-Net DL architectures (2DCNN, 3DCNN, ConvLSTM, Siamese, Siamese-Diff) utilizing features such as infrared (IR), water vapor (WV) channels, soil moisture (SM), elevation, and geographical coordinates. This approach effectively captures localized precipitation patterns across Italy and establishes a promising framework for future development of 1-km high-resolution global precipitation products
DTE Hydrology Evolution
Generate precipitation product- 1km precipitation over Mediterranean area
In this project, my responsibility is to developed a method to generate high-resolution (1 km) daily precipitation and rainfall maps over the Mediterranean area by combining multiple existing precipitation datasets. This method involves downscaling coarse resolution data from sources like CPC, GPM-LR, and SM2RAIN-ASCAT, guided by patterns derived from the CHELSA climate dataset. You can find the detail on this here.
See for precipitation product on the platform https://explorer.dte-hydro.adamplatform.eu/?use_case=1
Presentation project at Science Hub in ESA for
Advisory Committee for Earth Observation (ACEO)
OEMC
Develop a high resolution Flood Susceptibility Map (FSM).
In this project, my responsibility is to develop high resolution (1km) flood susceptibility maps. In this framework, the Random Forest machine learning algorithm was employed, and it was trained on a dataset of flooded and non-flooded areas (Global Flood Database v1). Six critical factors influencing flooding events use as input of Random Forest model with 1 km spatial resolution: elevation, slope, land cover, proximity to rivers, drainage density, and precipitation.
FSM framework
SM2RAIN-Climate
Monthly global long-term rainfall dataset for climatological studies
I develop SM2RAIN-Climate dataset which is a unique product which uses the European Space Agency (ESA) Climate Change Initiative (CCI) soil moisture v06.1 to provide monthly global rainfall for the 24-year period 1998–2021 at 1-degree spatial resolution. reliably estimates rainfall from satellite soil moisture data, offering a valuable independent dataset with strong global performance and potential for capturing rainfall trends.
SM2RAIN-Climate
Flood Forecasting framework
The title of my Ph.D. dissertation was “Flood Forecasting Framework under Semi-Arid Region.”
This study aims to develop a coupled hydrologic-hydraulic model for flood forecasting and warning system. In this study, the SWAT model as a rainfall-runoff generator is forced by the bias-corrected satellite precipitation products and replaces the routing scheme with the 2D hydraulic model (HEC-RAS) in order to predict localized flood depths, velocities, inundation map, and bankfull flow to set a threshold for the flood warning system. In addition, flood maps was extracted from Sentinel-1 Synthetic Aperture Radar (SAR) to assess the inundation map of couple model.
Study Area (Karkheh River Basin, Iran)
SWAT output
HECRAS output