Research
Current Research
Temperature Effects on Zoonotic Pathogens
Environmental Health Research Group, CULS Prague
Modelling how temperature variability influences the transmission and distribution of zoonotic pathogens using Distributed Lag Non-Linear Models (DLNM) and Generalized Additive Models (GAM). The work focuses on capturing delayed and non-linear exposure-response relationships across time and space.
Projects
Pesticide Spill Transport Simulation — DRUtES
Applied Hydropedology, CULS Prague · Co-authors: Fariba Naghizadeh, Maelia Joyeux
Numerical simulation of pesticide transport from an underground storage tank spill through layered vadose zone soil to a groundwater table 10m deep, set in the Příbram region, Czech Republic.
The simulation couples the Richards equation for variably saturated flow with the Advection-Dispersion Equation (ADE) for contaminant transport, implemented using the DRUtES finite element solver. A sensitivity analysis was conducted across multiple parameterisations of the unknown soil layer (gravel, sand, clay, fractured rock) to assess its influence on contaminant breakthrough curves and arrival times at the water table. Results were presented as an EGU-style scientific poster.
Tools: DRUtES, GMSH, Richards equation, ADE, Van Genuchten parameters
Cheminformatics Analysis of Cyanobacterial Metabolites
Case Western Reserve University, Cleveland · Supervisor: Dr. Mathew Bertin
Integrated cyanobacterial metabolites from the CyanoMetDB database using SMILES-based molecular representations. Assessed molecular similarity using fingerprint-based methods including Tanimoto, Tversky, MAP4, and GESim coefficients. Visualised chemical space clustering patterns using PaCMAP and Plotly.
Tools: Python, RDKit, Morgan fingerprints, PaCMAP, Plotly
Earth System Emulation using Machine Learning
BIT, Durg · Supervisor: Professor Sudip Bhattacharya
Evaluated machine learning and generative deep learning approaches for emulating Earth system model outputs using the ClimateBench v1.0 benchmark. Models assessed include Gaussian Process Regression, Support Vector Regression, Kernel Ridge Regression, Variational Autoencoders, and Generative Adversarial Networks. Results visualised using Cartopy, Iris, and Xarray.
Published findings in the International Journal of Computer Applications and as a conference paper on Gaussian Process Regression for climate modelling.
Tools: Python, GPR, VAE, GAN, Cartopy, Xarray
Regional Climate Modelling — Chhattisgarh, India
BIT, Durg
Downscaled CMIP6 data to enhance regional climate projections for Chhattisgarh, India. Incorporated physics-informed neural networks (PINNs) into spatial data pipelines for detecting and attributing weather extremes. Presented findings at the 19th Chhattisgarh Young Scientist Congress, NIT Raipur.
Tools: Python, PINNs, CMIP6, NetCDF
Large Network Data Analysis
IIT Indore · Supervisor: Dr. Aniruddha Singh Kushwaha
Parsed the Stanford Large Network Dataset Collection (Amazon co-purchasing networks) to extract features for machine learning applications. Built graph structures using NetworkX and the Stanford Network Analysis Platform (SNAP).
Tools: Python, NetworkX, SNAP