This tutorial shows how to create an interactive side-by-side map visualization of the mean temperature for 2024 and 2100 for the Chaudière-Appalaches region in Quebec using {leaflet} and {leaflet.extras2} R packages.
{pollen} and {vegperiod} are two R packages that can be used to analyze temperature, Growing Degree Days (GDD), and vegetation period. In this analysis, we explore historical temperature records, GDD trends, and vegetation period changes in Chaudières-Appalaches, Quebec, using these packages. By combining data visualization and exploratory data analysis (EDA) techniques, we uncover key patterns and anomalies that shed light on climate-driven changes in the region.
Building upon previous analyses and predictive modeling, I details the process of creating this visualization, including data preparation, disaggregation to daily levels, and kriging for enhanced spatial resolution. The post culminates in an animated map that illustrates precipitation trends and anomalies over time, providing valuable insights for climate analysis, agriculture, and water resource management.
In this phase of the analysis, we aim to model precipitation patterns in Centre-du-Québec using machine learning techniques, leveraging historical climate and environmental data. We will train an XGBoost models and predict precipitation trends. Model performance will be evaluated using cross-validation and regression metrics to determine the most effective approach.
Understanding long-term precipitation patterns is crucial for climate research, agriculture, and water resource management. In this post, we analyze 30 years of high-resolution precipitation data from the AgERA5 dataset, focusing on a single administrative region in Quebec. Using exploratory data analysis (EDA) techniques, we uncover trends, seasonal variations, and anomalies to gain deeper insights into precipitation dynamics
In this phase of the analysis, we aim to model precipitation patterns in the St. Lawrence Lowlands using machine learning techniques, leveraging historical climate and environmental data. We will compare Random Forest, XGBoost, and Mars models to assess their ability to capture complex relationships and predict precipitation trends. Model performance will be evaluated using cross-validation and regression metrics to determine the most effective approach.
Understanding long-term precipitation patterns is essential for climate research, agriculture, and water resource management. In this post, we analyze 30 years of precipitation data from the AgERA5 dataset for St. Lawrence Lowlands, using exploratory data analysis (EDA) techniques to uncover trends, seasonal variations, and anomalies.
This week we are exploring historical emissions data from Carbon Majors. They have complied a database of emissions data going back to 1854. In the first and second part I did some EDA and created a spatio-temporal machine learning model. In this part, I’m creating an animated vizualisation of the data including the prediction.
I will list here all the little snipset of code that I look up all the time.