Seasonal Fire Weather Index Analysis – Italy (2007–2024)

📌 OverviewThis repository contains all the necessary scripts and input data to reproduce the analysis of the <b>seasonal Fire Weather Index (FWI)</b> indicators for Italy between <b>2007–2024 (18 years)</b>.The workflow consists of <b>three R scripts</b> that calculate mean/median FWI values, burned area statistics, and relationships between fire danger indicators and observed burned areas.⚠️ Important Notes🔹 Dataset provenanceThe FWI dataset used here originates from the Copernicus Climate Data Store.This dataset is <b>no longer supported by the data providers</b>. Data and documentation are provided <i>as is</i>. Users are encouraged to consult the CDS Forum for any discussions or clarifications.The original <b>NetCDF (nc) files</b> contained:<b>Seasonal data</b>: June–September (fire season)<b>Indicator</b>: Fire Weather Index (FWI)<b>Scenario</b>: RCP2.6<b>Aggregation</b>: Multi-model mean caseThese NetCDF files were <b>reprojected using Climate Data Operators (CDO)</b> to a regular lon/lat grid before being included here.Reprojection grid specification:gridtype = lonlatxsize = 1000ysize = 460xfirst = -45xinc = 0.11yfirst = 22yinc = 0.11📂 Folder StructureAfter downloading, place the folder in the following path:C:/NERO2025/The folder should contain:C:/NERO2025/│── Script 01 C3S Analysis.R # Step 1 – FWI annual mean &amp; median│── Script 02 Burned Areas Per Region.R # Step 2 – Burned areas per ecoregion│── Script 03 Scatter Plots FWI vs Burned Area.R # Step 3 – FWI–burned area link│── IB_ITA_2007_2024_v_1.shp # Shapefile of burned areas│── subsez_ecoregioni_IT_proj.shp # Shapefile of Italian ecoregions│── other shapefile files (.dbf, .shx, etc.)▶️ Workflow<b>Step 1: Annual FWI statistics</b>Run <code><strong>Script 01 C3S Analysis.R</strong></code>Calculates <b>mean and median seasonal FWI values</b> per year for Italy and per ecoregion.Outputs:<code>FWI_Mean_Table.csv</code><code>FWI_Median_Table.csv</code><b>Step 2: Burned areas per ecoregion</b>Run <code><strong>Script 02 Burned Areas Per Region.R</strong></code>Intersects burned areas with Italian ecoregions.Calculates:Absolute burned area per year &amp; ecoregionRelative burned area (% of ecoregion burned)Performs <b>trend analysis</b> using linear regression &amp; Mann-Kendall.Outputs:<code>BurnedArea_Ecoregions.csv</code><code>RelativeBurnedArea_Ecoregions.csv</code><code>BurnedArea_Ecoregions_Trends.csv</code>Plots saved in <code>BurnedArea_Trends_Plots/</code><b>Step 3: Linking FWI and burned areas</b>Run <code><strong>Script 03 Scatter Plots FWI vs Burned Area.R</strong></code>Merges FWI tables (Step 1) with burned area statistics (Step 2).Generates <b>scatterplots with regression fits</b>, showing the relationship between FWI and burned areas.Outputs:Plots saved in <code>FWI_Burned_ScatterPlots/</code>CSVs of data used for each plot⚠️ <b>Dependencies</b>:<br><b>Script 03 </b>requires the results of <b>Scripts 01 &amp; 02</b>. Therefore, always run them in sequence.

📌 概览
本研究仓库包含复现2007年至2024年（共18年）意大利季节性火灾天气指数（Fire Weather Index, FWI）指标分析所需的全部脚本与输入数据。整体分析流程包含3个R语言脚本，分别用于计算FWI均值与中位数、过火面积统计量，以及火灾危险指标与实测过火面积之间的关联关系。

⚠️ 重要说明
🔹 数据集来源
本研究使用的FWI数据集源自哥白尼气候数据存储（Copernicus Climate Data Store）。该数据集已不再由数据提供商提供维护，本数据集与相关文档仅按"现状"提供，用户可前往CDS论坛（CDS Forum）进行讨论或咨询相关问题。

原始NetCDF（nc）文件包含以下内容：
- 季节数据：6月至9月（林火季）
- 指标：火灾天气指数（Fire Weather Index, FWI）
- 情景：RCP2.6
- 聚合方式：多模式平均情景

本仓库收录的NetCDF（nc）文件已通过气候数据运算符（Climate Data Operators, CDO）重投影至规则经纬度网格，重投影网格参数如下：
gridtype = lonlat（网格类型：经纬度网格）
xsize = 1000（X方向像素数：1000）
ysize = 460（Y方向像素数：460）
xfirst = -45（X起始坐标：-45）
xinc = 0.11（X方向步长：0.11）
yfirst = 22（Y起始坐标：22）
yinc = 0.11（Y方向步长：0.11）

📂 文件夹结构
下载完成后，请将该文件夹放置于以下路径：C:/NERO2025/
该文件夹应包含以下内容：
C:/NERO2025/
│── Script 01 C3S Analysis.R # 步骤1——FWI年度均值与中位数计算
│── Script 02 Burned Areas Per Region.R # 步骤2——各生态区过火面积统计
│── Script 03 Scatter Plots FWI vs Burned Area.R # 步骤3——FWI与过火面积关联分析
│── IB_ITA_2007_2024_v_1.shp # 过火面积矢量形状文件（Shapefile）
│── subsez_ecoregioni_IT_proj.shp # 意大利生态区矢量形状文件（Shapefile）
│── 其余矢量形状文件（.dbf、.shx等）

▶️ 分析流程
<b>步骤1：年度FWI统计量计算</b>
运行<code><strong>Script 01 C3S Analysis.R</strong></code>，可逐年计算意大利全国及各生态区的季节性FWI均值与中位数。
输出文件：
<code>FWI_Mean_Table.csv</code>
<code>FWI_Median_Table.csv</code>

<b>步骤2：各生态区过火面积统计</b>
运行<code><strong>Script 02 Burned Areas Per Region.R</strong></code>，该脚本将过火面积矢量与意大利生态区矢量进行空间相交分析，可计算：
1. 各年度、各生态区的绝对过火面积
2. 相对过火面积（占对应生态区总面积的百分比）
同时通过线性回归与Mann-Kendall检验开展趋势分析。
输出文件：
<code>BurnedArea_Ecoregions.csv</code>
<code>RelativeBurnedArea_Ecoregions.csv</code>
<code>BurnedArea_Ecoregions_Trends.csv</code>
生成的可视化图表将保存至<code>BurnedArea_Trends_Plots/</code>目录。

<b>步骤3：FWI与过火面积关联分析</b>
运行<code><strong>Script 03 Scatter Plots FWI vs Burned Area.R</strong></code>，该脚本将步骤1生成的FWI统计表格与步骤2的过火面积统计数据进行合并，生成带有拟合回归曲线的散点图，以展示FWI与过火面积之间的关联关系。
输出内容：
可视化图表保存至<code>FWI_Burned_ScatterPlots/</code>目录
各图表对应的原始数据CSV文件

⚠️ <b>运行依赖</b>：
<b>脚本03</b>依赖<b>脚本01与脚本02</b>的运行结果，因此请务必按顺序执行三个脚本。