Raj Rani

Scientific lead of BNP Paribas' global biodiversity (covering all countries) and ecosystem-services stress-testing framework under NGFS, ECB, EBA transition scenarios. Developing country-level projections to 2050, translating climate and socio-economic drivers into quantitative ESG and biodiversity risk indicators. Designed the end-to-end climate → biodiversity → ecosystem-service → economic-impact modelling chain. Developed ecosystem-specific transfer function/models (ecosystem services like: forest timber, CO₂ sequestration, cropland pollination, urban etc.) sensitive to climate extremes and long-term change Delivering Phase-1 scientific and technical outputs (methodology, R models, indicators, geospatial analysis) for integration into BNP’s stress-testing framework and defined the Phase-2 roadmap.

Designed and implemented climate-driven risk models under IPCC scenarios to quantify agricultural exposure, yield variability, and climate-related financial risk across Southern Europe. Developed a soil water balance and drought-risk modelling framework (FAO-56 based) using ERA5 reanalysis and CMIP5 (GCM–RCM / DRAIS) projections, integrating historical baselines with future climate pathways. Designed and implemented an analogue-based time-shift model to produce probabilistic soil-moisture projections through 2100, enhancing dynamic climate risk assessments and supporting robust yield prediction and financial risk evaluation. Quantified changes in the frequency and intensity of extreme and compound events (droughts, heatwaves, heavy precipitation) across near-, mid-, and long-term horizons under IPCC RCP scenarios, translating climate hazards into actionable financial risk indicators. Contributed to the development of operational climate-resilience and risk-assessment tools, supporting long-term financial risk management and adaptation strategies in the agricultural sector. skills : Python · ERA5 · CMIP5 (GCM–RCM, DRAIS) · Extreme Value Theory · Probabilistic Forecasting · Soil Water Balance Modelling · Time-Series & Quantile Analysis · IPCC Scenarios · Data Visualization ·Data Visualization· Principal Component Analysis · Probabilistic Forecasting

DataWave: Collaborative Gravity Wave Research (Virtual Earth System Research Institute (VESRI), Schmidt Futures) funded by google • Parameterized gravity wave drag schemes in LMDZ-IPSL model using ERA5 data to improve Quasi-Biennial Oscillation (QBO) simulations—critical for seasonal forecasting and extreme event prediction • Conducted statistical validation comparing CMIP6 model outputs with stratospheric balloon observations (Strateole-2, Google Loon) to reduce stratospheric uncertainty • Contribute to optimized gravity wave parameterizations using data assimilation methods including ensemble Kalman filtering and stochastic optimization in CMIP7 framework. • Advanced understanding of middle atmosphere dynamics and its influence on long-term climate predictions Tech: Python, Fortran, shell scripting, LMDZ, CMIP6/7, ERA5, Data Assimilation, High-Performance Computing (HPC), middle atmospheric dynamics, Data visualization

• Analyzed climate system variability during Last Glacial Maximum using MPI Earth System Model variability across ocean, atmosphere, and vegetation systems under different physics and boundary conditions during Last Glacial Maximum. • Statistical analysis and validated model performance against paleoclimate proxy data, identifying biases and improving model reliability for future projections. • Assessed climate sensitivity by comparing simulations with different ice-sheet configurations, providing insights into system stability. Tech: MPI-ESM, Python, R, shell scripting, Fortran, Climate Data Analysis, Statistical analysis, Paleoclimate.

-Yvette, Île-de-France, France • Reconstructed Atlantic Ocean water mass evolution over 40,000 years using particle filter data assimilation in the iLOVECLIM Earth System Model, improving accuracy of past climate benchmarks. • Quantified climate change and CO2 impacts on global vegetation patterns using CMIP5/PMIP3 frameworks, identifying key risk drivers for global vegetation across past and present climates- focus on LGM, Mid-Holocene and pre-industrial. • Led coupling of BIOME4 vegetation model with climate models and statistical validated results against pollen proxy data, enhancing predictive capability for vegetation-climate dynamics. Tech: Python, R, Fortran, CDO, shell scripting, iLOVECLIM, BIOME4, CMIP5/6/PMIP3, Particle Filter Data Assimilation, Statistical Analysis, vegetation modelling, Data Visualization

• Reconstructed Holocene climate patterns across the Euro-Mediterranean region using data assimilation, Statistical downscaling techniques, Earth system model and vegetation model. • Integrated pollen-based vegetation models to improve regional climate understanding and inform future climate risk projections • Applied particle filter methods and LOVECLIM modeling to enhance paleoclimate accuracy in a climatically vulnerable region Tech: Climate Reconstruction · Data Assimilation · Statistical Downscaling · LOVECLIM · Particle Filter · Paleoclimate Analysis · Fortran · R

• Developed data assimilation techniques for identification of unknown atmospheric emissions • Applied inverse modelling methods to atmospheric pollution tracking Tech: Inverse Modelling · Atmospheric Emissions · Data Assimilation · Air Quality · Environmental Monitoring, Fortran and MATLAB, shell scripting