Teacher: Giovanna Venuti
Tutor: Dr. Daniela Stroppiana – CNR/IREA
Description: This project focuses on the construction of a harmonised, multi-source dataset to analyse the impact of wildfires on key lake water quality parameters, specifically chlorophyll-a and turbidity. Two main tasks should be addressed:
- Data preparation and pre-processing
- Exploratory Analysis
The first stage involves building a consistent datasets at the lake catchment level:
- Import and merge multi-source datasets (fire data, meteorological variables, lake properties, anthropogenic indicators)
- Align spatial and temporal resolutions
- Handle missing values and outliers
- Apply normalization/scaling and transformations (e.g., log-transformations)
- Generate lagged variables (e.g., precipitation, runoff, fire metrics)
- Structure data into time series format
The second stage will focus on the development of tools/scripts to:
- Compute correlation matrices and visualize relationships
- Implement Generalized Additive Models (GAMs) to explore nonlinear effects
- Produce diagnostic plots (smooth terms, residuals)
- Evaluate multicollinearity (e.g., VIF) and filter predictors
The resulting dataset is used to support the implementation of statistical models aimed at quantifying the influence of environmental and anthropogenic drivers on post-fire water quality dynamics. All datasets should be available at the global scale and for the period 2017-2022.