aymen-hadj-mebarek/Algeria-climate-analysis

# Algeria Climate Analysis and Prediction This project is divided into two phases, focusing on **data preparation** and **machine learning applications** using a climate dataset covering **Algeria**. The goal is to transform raw, heterogeneous data into actionable insights through systematic preprocessing, analysis, and predictive modeling. --- ## Collaborators | | | |:---------------------------------------------------------------------------------------------:|:---------------------------------------------------------------------------------------:| | **Aymen HM.** | **Yasmine A.** | ### **Phase 1: Data Preprocessing** #### **Data Sources** - **Soil-DATA** : which contains soil properties of the Algerian territory.\ *you can find more infos about this data set here (for this one go to "Soil Attributes per depth layer")*\ **due to the size of the dataset, i couldnt upload it to Github. You can download the dataset using the api to have the exact data set i used from here** - **Climate-DATA** : which contains climate variables of 2019 for the entire world.\ *you can find more infos about this data set here* - **Country-DATA** : which contains country varuiables of the entire world. #### **Dataset Overview** - **Raw Data Format**: Climate and soil data stored in `.nc` (NetCDF) files, a format commonly used for multidimensional geospatial datasets (e.g., latitude, longitude, time). - **Initial Data Shape**: - **Global Climate Data**: Dimensions of `(time=365, lat=1800, lon=3600)` with variables like temperature, precipitation, and humidity. - **Soil Data**: Similar spatial resolution but includes variables like soil moisture, pH, and texture. --- #### **Key Preprocessing Steps** 1. **Extracting Algeria’s Data** - **Geospatial Filtering**: - Used a georeferenced Algeria boundary map (shapefile) to mask global NetCDF datasets. - Applied spatial indexing to extrac …