Data Downloading and Processing

Step 1: Download SAR Data

Adjusting the process to download SAR data from Sentinel-1, ALOS-PALSAR, Radarsat-1, or
other preferred satellites:

Go to ASF Data Search (https://search.asf.alaska.edu/).

Use the map interface to define your Area of Interest (AOI).
Filter by satellite (e.g., Sentinel-1, ALOS-PALSAR), polarization, and other parameters.
Download the files in GeoTIFF format.
Extract and store them in a working directory for ArcGIS.

Step 2: Preprocessing of SAR Data

Perform equivalent preprocessing in ArcGIS using Geoprocessing Tools:

Mosaic Rasters (if needed):

Use the Mosaic to New Raster tool to combine multiple tiles into one raster.
Input: SAR images.
Output: A single, seamless raster.
Radiometric and Terrain Corrections:
If required, apply radiometric or terrain corrections using external tools like SNAP, then import
corrected files into ArcGIS.

Step 3: Layer Preparation

Add SAR layers to the Table of Contents in ArcMap.
Use Composite Bands (if working with multiple bands):
Go to Data Management Tools > Raster > Composite Bands.
Input: Relevant SAR bands (if applicable).
Output: A composite raster.

Step 4: Region of Interest (ROI) Creation

Use the Editor Toolbar:
Start an editing session and draw polygons over areas of interest.
Assign attributes like class names (e.g., water, vegetation) in the attribute table.
Save the ROIs as a shapefile.

Step 5: Spectral Signature Creation

Open the Training Sample Manager (part of the Image Classification Toolbar in ArcGIS).
Add training samples (ROIs) by selecting polygons from your ROI shapefile.
Assign class names or IDs to each sample.
Export spectral signatures as a signature file (.gsg).

Step 6: Image Classification

Open the Image Classification Toolbar:
Go to Customize > Toolbars > Image Classification.
Select the Maximum Likelihood Classification tool:
Input: Preprocessed SAR raster and .gsg file.
Output: Classified raster.

Step 7: Post-Classification Processing

Refine the classified raster:

Use Reclassify (Spatial Analyst Tools):

Input: Classified raster.
Reassign or simplify class IDs.
Apply a Majority Filter to smooth classification results:
Go to Spatial Analyst Tools > Generalization > Majority Filter.
Input: Classified raster.

Step 8: Accuracy Assessment

Compare the classified raster against ROIs:
Use the Confusion Matrix tool in the Spatial Analyst extension.
Generate an error matrix and calculate overall accuracy.

Step 9: Export Results

Save the classified raster in GeoTIFF or other preferred formats using Data Export.
Create a map layout:
Use the Layout View in ArcGIS to design maps.
Add legends, titles, and other elements to match the QGIS output.
General Workflow for Tutorial Tasks

Conversion of DN to Reflectance (if applicable):

Use ArcGIS Raster Calculator with reflectance formulas based on metadata (e.g., solar
elevation, gain, and bias values).
Alternatively, preprocess data outside ArcGIS (e.g., SNAP for Sentinel-1) and import
corrected files.
ROI Creation and Training Data Preparation:

Use the Editor Toolbar to digitize polygons over ROIs.

Use the Training Sample Manager to assign spectral class names or IDs.
Classification:

Leverage Image Classification Toolbar and algorithms like Maximum Likelihood for
supervised classification.
For unsupervised classification, use Iso Cluster Unsupervised Classification (Spatial Analyst
Tools).
Accuracy Assessment:

Generate confusion matrices using the Confusion Matrix tool to assess classification quality.
1. Land Cover Classification (e.g., Vegetation, Soil, Built-up, Water)

Step 1: Load the SAR Data

Input: SAR data from Sentinel-1, ALOS-PALSAR, or Radarsat-1 in GeoTIFF format.
Tool: Add Data (ArcMap).
Action: Import all SAR bands.

Step 2: Preprocessing the Data (Radiometric Correction)

If radiometric corrections are done externally (e.g., using SNAP or Sentinel Toolbox), simply
import the corrected SAR data into ArcGIS.
Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator) for custom
correction formulas based on metadata (e.g., scaling factor, radiometric calibration).
Formula: Apply the calibration formula specific to the dataset, such as:
makefile
Copy code
SAR_output = SAR_raw * Calibration_Factor
Output: Corrected SAR raster (corrected_sar.tif).

Step 3: Create a Virtual Raster (Band Composite)

Tool: Composite Bands (Data Management Tools > Raster > Raster Processing).
Inputs: Select the SAR bands (e.g., VV, VH).
Output: Composite raster (composite_raster.tif).

Step 4: Create Training Data (ROIs)

Tool: Create Features (Editor Toolbar).
Action:
Start an editing session.
Use the Polygon Tool to draw polygons for each land cover class (vegetation, soil, water,
built-up).
Output: ROI shapefile (roi_data.shp).

Step 5: Extract Spectral Signatures for Training

Tool: Training Sample Manager (Image Classification Toolbar).
Input:
Training Data: The ROI shapefile (roi_data.shp).
Raster Input: The composite raster (composite_raster.tif).
Output: Signature file (signature_file.gsg) containing spectral signatures for each class.

Step 6: Perform Supervised Classification

Tool: Maximum Likelihood Classification (Image Classification Toolbar).
Input:
Raster Input: Composite raster (composite_raster.tif).
Signature File: Spectral signature file (signature_file.gsg).
Output: Classified raster (classified_output.tif).

Step 7: Post-Processing the Classification

Tool: Reclassify (Spatial Analyst Tools > Reclass).
Input: Classified raster (classified_output.tif).
Reclassification Values:
For example, reassign class IDs: water = 1, vegetation = 2, built-up = 3, soil = 4.
Output: Reclassified land cover raster (reclassified_land_cover.tif).
2. Flood Monitoring

Step 1: Calculate NDVI for Water Detection

Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator).
Formula: (NIR - Red) / (NIR + Red) to calculate NDVI.
Input:
NIR Band and Red Band from Sentinel-2 or Landsat imagery.
Output: NDVI raster (ndvi.tif).

Step 2: Threshold for Water Masking

Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator).
Formula: Con("ndvi.tif" < 0, 1, 0) to detect water areas (NDVI < 0).
Input: NDVI raster (ndvi.tif).
Output: Water mask (water_mask.tif).

Step 3: Post-Flood Assessment

Tool: Raster Difference (Spatial Analyst Tools > Math > Minus).
Input:
Pre-flood NDVI raster and Post-flood NDVI raster.
Output: Difference raster (flood_difference.tif).
3. Wildfire Monitoring (Burned Area Detection)

Step 1: Calculate NBR (Normalized Burn Ratio)

Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator).
Formula: (NIR - SWIR) / (NIR + SWIR) for NBR calculation.
Input: NIR and SWIR bands from Landsat or Sentinel-2 imagery.
Output: NBR raster (nbr.tif).

Step 2: Calculate dNBR (Change in NBR)

Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator).
Formula: PreNBR - PostNBR to calculate the difference in NBR (dNBR).
Inputs: Pre- and post-fire NBR rasters.
Output: dNBR raster (dnbr.tif).

Step 3: Identify Burn Severity

Tool: Reclassify (Spatial Analyst Tools > Reclass).
Input: dNBR raster (dnbr.tif).
Reclassification Values:
Low burn severity: dNBR < 0.1.
Moderate burn severity: dNBR between 0.1 and 0.3.
High burn severity: dNBR > 0.3.
Output: Burn severity raster (burn_severity.tif).
4. Forest Monitoring (Land Cover Change Detection)

Step 1: Perform Classification on Pre- and Post-Images

Follow the same Land Cover Classification workflow from Steps 1-6 for both pre- and post-
classification images.
Inputs: Pre- and post-classification images.
Outputs: Two classified rasters, one for each time period.

Step 2: Calculate Land Cover Change

Tool: Raster Difference (Spatial Analyst Tools > Math > Minus).
Input: Pre- and post-classified rasters.
Output: Land cover change raster (land_cover_change.tif).
5. Surface Temperature Estimation

Step 1: Convert DN to Reflectance and Temperature

Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator).
Formula for conversion to Brightness Temperature: TB = K2 / ln((K1 / Radiance) + 1).
Inputs: Landsat thermal bands (Band 10, Band 11).
Output: Brightness Temperature raster (brightness_temp.tif).

Step 2: Reclassify for Emissivity

Tool: Reclassify (Spatial Analyst Tools > Reclass).
Input: Classified raster.
Reclassification Values: Assign emissivity values based on land cover class:
Water: 0.98.
Vegetation: 0.98.
Built-up: 0.94.
Soil: 0.93.
Output: Emissivity raster (emissivity.tif).

Step 3: Calculate Land Surface Temperature (LST)

Tool: Raster Calculator (Spatial Analyst Tools > Map Algebra > Raster Calculator).
Formula: LST = TB / (1 + (λ * TB / c2) * ln(emissivity)), where λ is the wavelength of the
emitted radiance and c2 is a constant.
Inputs: Brightness temperature raster (brightness_temp.tif) and emissivity raster
(emissivity.tif).
Output: LST raster (lst.tif).

6. Final Outputs and Map Layouts

Tool: Layout View (File > New > Layout).
Add classified rasters, legends, scale bars, titles, and other elements to the map.
Export: Save the final map layout as PDF or Image.