 IIT Mandi AI-Hydrology Lab |
 20 peer-reviewed journals · books · conferences |
 50+ via Google Scholar profile → |
 India · Himalayas South-Asian basins |
class SiddikBarbhuiya:
role = "PhD Researcher, AI-Hydrology"
affiliation = "Indian Institute of Technology Mandi"
research = ["Differentiable Hydrology",
"Deep Learning Rainfall–Runoff Modelling",
"Climate Change Impacts & Extremes",
"Geospatial / Remote Sensing"]
stack = ["PyTorch", "JAX", "xarray", "GEE",
"QGIS", "SWAT", "GR4J", "VIC", "NeuralHydrology"]
currently = "Scaling differentiable land-surface models for India"
open_to = ["Research Collaborations", "Open-Source Hydrology",
"PhD / Postdoc Discussions"]
def say_hi(self):
print("Let's collaborate on water + AI → siddikbarbhuiya@gmail.com")|
Hybrid physics-ML models (MILC-δ, GAN-PO) embedding neural networks inside differentiable physics frameworks for distributed land-surface modelling. Rainfall–runoff modelling across 144+ Indian basins — GRU, LSTM, EA-LSTM, AR-LSTM, Transformer, Informer, Reformer, Linformer, plus state-space (Mamba) experiments. |
Compound extremes, ETCCDI indices, CMIP6 streamflow projections across South Asia and the Himalayas; nonstationary FFA. Cloudburst, landslide, flood risk with satellite + gauge fusion (ERA5, IMD, SAR), WRF dynamics, and remote-sensing soil-moisture ML. |
📌 Latest: "Differentiable, Learnable MILC: Balancing Predictive Skill and Physical Interpretability" — EGU 2026 📌 2026 Journal: Hydrological Processes 40(3) e70476 — CMIP6 streamflow projections for India 📌 2025 Journal: Environmental Modelling & Software 194 — Gauged-to-Ungauged Deep Learning RR Modelling
Most-cited & most-recent (click to expand all 20 →)
| Year | Title | Venue | Citations |
|---|---|---|---|
| 2026 | Hydro-meteorological & Infrastructural Damage Analysis of the Ramban Cloudburst (J&K) | Int. J. of Disaster Risk Reduction (Elsevier) | — |
| 2026 | Projections of Future Streamflow for India Informed by CMIP6 GCMs | Hydrological Processes 40(3) e70476 | 1 |
| 2026 | Differentiable, Learnable MILC | EGU 2026 | — |
| 2025 | From Gauged to Ungauged: Large-Scale DL Rainfall-Runoff for India | Env. Modelling & Software 194 (Elsevier) | 3 |
| 2025 | Climate-Change Impacts on Beas River Basin Hydroclimate | Springer book chapter | 2 |
| 2024 | Performance Evaluation of ML in Hydrology: SWAT vs GR4J vs DL | J. Earth System Science 133(3) 136 | 15 |
| 2023 | Streamflow in Ungauged Basins via Physical Similarity | Arabian J. of Geosciences 16(12) 672 | 7 |
| 2023 | Nonstationary Flood Frequency Analysis: Review of Methods & Models | Springer book chapter | 12 |
| 2021 | Lumped Conceptual Rainfall-Runoff Genie Rural Models for Streamflow | Springer (HYDRO 2021) | 6 |
Plus 11 conference papers at AGU Fall Meetings (2023–2025) and EGU General Assemblies (2023–2026).
|
Notebooks & scripts for hydrological analysis — preprocessing, calibration, evaluation.
|
AI for Disaster Risk Management — flood, landslide and cloudburst analytics.
|
|
Geospatial & hydrologic data viz recipes — Matplotlib, Cartopy, Plotly.
|
Source for barbhuiya12.github.io/Portfolio — React + TypeScript + Vite.
|
Side things I tinker with after-hours — web stuff for the research community I'm part of.
|
International Conference on Climate, Disaster Risk & Hydrology
|
Indian Climate Information Explorer
|
"Towards models that respect the physics of water — and learn from its data."