Improving Land Surface Temperature (LST) Maps Using Night-time Images

Heatwaves_NewDataSource/README.md

@@ -0,0 +1,157 @@
+**Improving Land Surface Temperature (LST) Maps Using Nighttime Images**
+
+Land Surface Temperature (LST) is a key indicator for analyzing
+heatwaves and the urban heat island (UHI) effect, which refers to higher
+surface and air temperatures in urban areas compared to their rural
+surroundings.
+
+In the CLIMAAX Heatwave Risk Assessment Workflow, satellite-based LST
+images as the exposure component is analysed. The workflow currently
+defines the Landsat 8 image downloading methodology and related codes.
+
+At the **Şanlıurfa and Mersin Metropolitan Municipalites**, Landsat 8
+datasets for daytime observations during the summer season have been
+already applied. However, since daytime land surface temperatures often
+exceed 60°C and are distributed relatively homogeneously across the city
+center, it was not possible to effectively visualize temperature
+variations or the urban heat island effect.
+
+Therefore, as joint working group from Mersin and Şanlıurfa Metropolitan
+Municipalities, we searched for another data source providing nighttime
+LST images. Two of the most widely used satellite sources for LST data
+are Terra ASTER (Advanced Spaceborne Thermal Emission and Reflection
+Radiometer) and Landsat 8 (Thermal Infrared Sensor – TIRS).
+
+Below is a comparison of their sensor and data characteristics:
+
+| **Feature** | **Terra ASTER** | **Landsat 8 (TIRS)** |
+|----|----|----|
+| Satellite Platform | Terra (NASA, launched 1999) | Landsat 8 (NASA/USGS, launched 2013) |
+| Spatial Resolution | 90 meters | 30 or 100 meters |
+| Day/Night Imagery | Day or Night | Daytime only |
+| Data format | GeoTiff/HDF | GeoTiff |
+| LST Unit | Kelvin | Degrees Celcius |
+| Data-Source | NASA Earthdata Search | RSLab Landsat LST |
+| URL | <https://search.earthdata.nasa.gov/> | <https://rslab.gr/Landsat_LST.html> |
+
+**Why Terra ASTER Instead of Landsat 8?**
+
+Landsat 8 offers consistent global coverage and good temporal
+consistency but has fewer thermal bands and generally only provides
+daytime acquisitions. For city-scale UHI studies, ASTER’s 90 m thermal
+data is better suited to detect local temperature variations between
+urban structures and vegetated areas. ASTER’s ability to acquire both
+daytime and nighttime thermal images is a major advantage. Nighttime LST
+helps capture residual heat storage in built-up areas and provides
+clearer contrasts between urban and rural surfaces. This makes ASTER
+data more convenient and reliable for visualizing and quantifying the
+urban heat island effect, especially in densely built city centers.
+
+**Main disadvantages of ASTER Data**
+
+- Data analysis is more complex than for Landsat 8.
+
+- A simple additional code is required to convert the LST unit from
+  Kelvin to degrees Celsius.
+
+- Preloaded “granules”should be clipped for desired area of concern.
+
+In Şanlıurfa, we compared both data sources over the same location using
+ASTER night-time and Landsat 8 daytime maps. The results clearly
+demonstrated that night-time ASTER imagery provides a more effective
+visualization of the urban heat island effect.
+
+<img src="./media/image1.png" style="width:6.3in;height:2.28351in" />
+
+We have improved the Climaax Heatwave Risk Map of Şanlıurfa MM and
+Mersin MM city centres by applying ASTER’s night-time LST data-sets:
+
+<img src="./media/image2.png" style="width:6.29565in;height:2.5424in" />
+
+<img src="./media/image3.png"
+style="width:6.25865in;height:2.48391in" />
+
+**Step by Step Data Processing (We applied)**
+
+1\. Go to : <https://search.earthdata.nasa.gov/>
+
+2\. Sign in.
+
+3\. Search data source “ASTER Surface Kinetic Temperature”
+
+4\. Select data source “ASTER L2 Surface Kinetic Temperature V003”
+
+5\. “Spatial” function: Navigate in the map, locate the Region and draw
+a polygon or circle.
+
+6\. “Temporal” function: Define the temporal period as 01.07 to 31.08 of
+e.g. 2020-2025 (tick «Use a recurring date range» box)
+
+7\. Select «Night»
+
+8\. Select the best preloaded «granule» covering our region. Probably
+much bigger than our area of interest. Select «GeoTiff» as file format.
+“Download” the selected granule (image). Requested image data will be
+sent via e-mail within a day
+
+9\. The incoming ASTER image file cannot be used directly in the
+workflow.
+
+The following pre-processing steps are required:
+
+- Convert the temperature values from Kelvin to Celsius.
+
+- Since the file covers a very large area, it needs to be clipped to the
+  study area.
+
+The software we used for this study is ArcMAP version 10.8. It is a
+desktop GIS software developed by Esri. This screen shows the appearance
+of the Aster satellite image within the software.
+
+10\. Since the surface temperature of the downloaded satellite image is
+in Kelvin, we need to convert it to Celsius. To do this, we use the
+Raster Calculator tool. After opening the tool, in the calculation field
+we enter:
+
+```("file_name.tif" * 0.1) - 273.15```
+
+and then select the folder where the new file will be saved.
+
+11\. Since the obtained image is very large, we need to cut it according
+to our study area. To do this, we created a shapefile (shp) centered on
+our study area. The square visible on the screen is centered on
+Şanlıurfa. In ArcMap, we open the Image Analysis tool. we select our
+study area. In the Image Analysis tool, we choose the image we want to
+clip (After conversion). Then, in the Processing section, the desired
+area is clipped.
+
+12\. Change the coordinate system of the clipped image. The reason for
+this step is to align it with the coordinate system of the population
+data in the workflow, ensuring that the workflow functions correctly. To
+do this, we used the Project Raster tool to change the image’s
+coordinate system. The output coordinate system is GCS_WGS_1984.
+
+13\. If the image name is incorrect, it causes an error in the workflow.
+Therefore, the image name needs to be changed. All files should be saved
+under the LST Folder.
+
+14\. The file — the updated ASTER image — is now ready and working
+properly in the workflow.
+
+**Conclusion**
+
+The results clearly demonstrated that night-time ASTER imagery provides
+a more effective visualization of the urban heat island effect. But,
+data analysis is more complex than for Landsat 8.
+
+We hope that, temperature conversion from Kelvin to Celsius and clipping
+operation can be performed within in the workflow by means of additional
+codes instead of manual ArcGIS operations.
+
+**For further explanation and contact:**
+
+**İhsan Topallı- MersinMM:** <ihsan@topalli-ai.com>
+
+**İzzettin Karabulut-Sanliurfa MM:** <karabulut6363@gmail.com>
+
+**Tamer Atalay- Atalay Climate Consulting:** tamer@atalayconsulting.com