Tools & Skills

Computational tools, programming languages, models, and platforms used in my research.

Programming & AI Frameworks

Languages

Python (NumPy, SciPy, xarray, Pandas, Dask), C++, Fortran, MATLAB, IDL, NCL.

Machine Learning & Deep Learning

PyTorch, TensorFlow (including Keras), JAX, scikit-learn, XGBoost, MXNet. Architectures: CNNs and residual networks, RNNs (LSTM, GRU), ConvLSTM, autoencoders, U-Net and DeepLabV3+ segmentation, GANs, and Transformer-based models. Physics-informed and physics-constrained neural networks for retrieval and parameterization problems.

Bayesian Inference & Uncertainty Quantification

Probabilistic programming with NumPyro and PyMC (MCMC/NUTS, posterior-predictive checks), bootstrap and ensemble methods, and large-ensemble Monte Carlo inverse modeling.

Cloud & HPC

Experience deploying and managing workloads on AWS and Azure cloud platforms. Proficient in high-performance computing (HPC) environments across Linux, macOS, and Windows; GPU training workflows.

Numerical Models

Climate & Weather Models

Community Earth System Model Version 2 (CESM2), E3SM, superparameterized CAM (SPCAM), WRF and WRF-Chem, CMIP6 multi-model ensemble analysis; single-column model configurations for process studies.

Radiative Transfer

libRadTran (DISORT solver), RRTM/RRTMG. Global, seasonally resolved clear-sky and cloudy-sky simulations of atmospheric absorptivity and in-cloud radiative heating.

Aerosol–Cloud & Dust Models

Developed and implemented cloud microphysics algorithms, including process-tagging tracers and secondary-ice mechanisms, within the 3-D Aerosol–Cloud (AC) Model (Fortran; Lund University). Large-ensemble inverse modeling for global dust cycle reconstruction; ERA5-driven dust emission modeling; DustCOMM observationally constrained dust products.

Satellite, Airborne & Observational Datasets

Multi-Sensor Satellite Data

CloudSat and CALIPSO (2B/L2 products), MODIS and VIIRS (L1B and products), IASI, CERES, PREFIRE, Himawari-8/9 AHI and GOES-16 ABI (L1 radiances), SEVIRI, MIDAS–Aqua/Terra, Megha-Tropiques ScaRaB/3, SAGE III/ISS, GPM/IMERG, and the CCIC ice-water-path climate record.

Airborne Field Campaigns

NASA airborne datasets from CPEX-CV, NAMMA, and GRIP (dropsondes, in-situ aerosol and cloud probes, and airborne radar/lidar) for simulation initialization and evaluation.

Ground-Based & In Situ

AERONET dust-dominated sites, HadISD surface-station winds, ARM North Slope of Alaska (NSA) observatory, PINE chamber INP measurements, aircraft in-situ probes (CIP, 2D-S), and field-campaign validation data.

Reanalysis Products

ERA5, MERRA-2, and three aerosol reanalyses for dust cycle constraint and model evaluation.