Malaria is a deadly disease caused by parasites that are transmitted to humans by mosquito bites from infected female Anopheles mosquitos. Malaria is most common in tropical and subtropical climates, where the parasites thrive.
According to World Health Organization (WHO), the estimated number of deaths from malaria stood at 409000 people in 2019, and the WHO African Region carried a disproportionately high share of the global malaria burden (constituted 94 percent of malaria cases and deaths).
This tutorial contributes to the fight against malaria by making use of Spatial Data Science and Remote Sensing techniques (ArcGIS Pro) to estimate areas that are prone to malaria using ArcGIS Pro. Many factors affect the prevalence of malaria in a Region. However, for the purposes of this tutorial, the following factors are considered; Normalized Difference Vegetation Index (NDVI), Land Use, Slope, and Distance to River. Figure 1 shows a summary of the methodology employed in this tutorial.
Estimation of NDVI from Landsat 7 Imagery
The Normalized Difference Vegetation Index (NDVI) can be downloaded freely from the United States Geological Survey (USGS) website. NDVI can also be calculated from Landsat 7 imagery using the raster calculator or the Landsat toolbox after performing line correction on the imagery. In order to correct for scanline errors, this video may be helpful (https://www.youtube.com/watch?v=kJGkl_Q7D3w).
The Landsat toolbox can be downloaded here (https://drive.google.com/file/d/1FKc-G1vMVtWXi66hh15VKoj-zk3UPWgX/view).
After scanline correction, calculation of NDVI can be done either from the Landsat toolbox or from the raster calculator.
To calculate NDVI using the Landsat toolbox, the following steps can be followed;
- Select “NDVI” from the “Landsat Toolbox,” which can be found as a part of the Arctoolbox.
- Set “red band” to be band 3 free of scanline errors.
- Set “Near IR band” to be band 4 free of scanline errors.
- Set “output raster” to be the desired output name and location of the raster to be saved.
To calculate NDVI using the raster calculator, the following steps can be followed;
- Search “raster calculator” from the geoprocessing search bar
- Input the following as shown in the figure below into the calculation bar of the raster calculator.
3. Click “Run.”
Preparing “Distance to River” Data
In order to estimate a set distance to river as an area susceptible to the breeding of mosquitoes, and therefore susceptible to malaria, hydrology or river data will be extracted from Digital Elevation Model (DEM) by the following steps:
- Select “Add Data” icon to import DEM data onto the map canvas.
- Type and search “fill” from the search bar of the geoprocessing toolbox.
- From the “fill” dialogue, set “input raster” as the DEM, and “output raster” as desired name and location of the file to be saved.
- Click “Run.”
- Type and search “flow direction” from the search bar of the geoprocessing toolbox.
- Set “Input Surface Raster (ISR)” to be the results of “fill,” and “Output Surface Raster (OSR) to be desired name and location. Click “Run.”
- Type and search “flow accumulation.”
- Set “Input raster” to be the results of “flow direction,” and set “output raster.”
- Click “Run.”
- Type and search “raster calculator,” and input the formula as shown below into the calculation bar of the window. Click “Run.”
- Type and search “stream order.” Set “input raster” to be the results of the raster calculator, and set “output flow direction raster” to be the resulting raster of flow direction. Set “output raster,” and “Run.”
- Type and search “stream to feature.” Set “input raster” to be the resulting stream order raster. Set “input flow direction raster,” and click “Run.”
Preparing the Slope Data
Slope is estimated from the DEM data by following the steps outlined below:
- Type and search “slope” from the search bar of the geoprocessing toolbox. The slope should be selected from the Spatial Analyst toolbox tab.
- Set “input raster” to be the DEM file, and “output raster” to be the desired name and location of the raster data to be saved.
- Set “output measurement” to be degree.
Preparing the Landsat Data
Land use data will be downloaded from the WaterItech website.
- Open wateritech.com, and select Data from the tab menu.
- Scroll down to Landcover map, and click to download. The downloaded file is archived and can be extracted into a folder.
- After, the map can be imported onto the map canvas of ArcGIS Pro by using “Add Data.”
- For the purposes of this tutorial, the WGS1984 UTM Zone 30N spatial reference system will be used for all the datasets since the study focuses on a section in Ghana.
- Land cover map will be projected from WGS 1984 to WGS 1984 Zone 30N. Type and search “Project Raster” from the search bar of the geoprocessing toolbox.
- Set “input raster” to be the downloaded Landcover map, and “Output coordinate system” to be WGS1984 UTM Zone 30N.
- Click “Run.”
“Reclassify” and Weighted Overlay Analysis
“Reclassify” will be used to standardize the data based on a set of defined classes. Weighted overlay analysis will also enable preferences to be given to the individual factors based on their level of importance. The following steps will be undertaken in performing “reclassify:”
- Type and search “reclassify” from the search bar of the geoprocessing toolbox. The “reclassify” tool should be selected from the Spatial Analyst toolbox.
- Set “input raster” to be results of the Land Cover map.
3. Select “classify,” and set the number of classes to be 5.
4. Click “Run.”
5. Repeat steps for the results of slope, Euclidean distance_rivers, and NDVI.
The following steps will be followed to perform weighted overlay analysis over the factors:
- Type and search “weighted overlay” from the search bar of the geoprocessing toolbox.
- From the drop-down menu on the rasters tab, select all the reclassified images of the factors; distance to rivers, land cover, slope, and NDVI.
- The following percentage values should be allocated to the factors as shown in figure below.
4. Click “Run.”