How can you map vegetation cover in GIS?

You need to decide the scale of vegetation cover mapping. If you want a very high resolution, you need to acquire it using UAV or commerical VHR satellite. If you are okay with a lower resolution, public satellite such as Sentinel, Landsat, and MODIS are good enough.

Acquiring the data differs across the different platforms. When I first started my GIS journey, I had to download GBs of raw LandSat imageries from earth explorer by NASA, ingest it into local client app (mostly QGIS/ArcGIS) and then run raster calculator. However, Google Earth Engine already has some of these raster calculations processed on the server side (their cloud compute) with the raw imageries not needed to be downloaded anymore.

Be careful here. A satellite image is the capture of a moment in time and with vegetation that can change rapidly. This will only work efficiently in small areas as you need to teach the computer what colour bands are what vegetation and if you have multiple adjacent images, perhaps captured on different dates, your analysis of the first image will not transpose over onto other images. I learnt this through experience and it takes a lot of time to correlate all the data into one coherent set of vegetation.

To map vegetation cover in GIS: 1. Collect: Obtain satellite/aerial imagery. 2. Preprocess: Correct image distortions. 3. Calculate: Use NDVI for vegetation health. 4. Classify: Categorize land into vegetation types. 5. Validate: Check accuracy with ground truth. 6. Visualize: Create and analyze vegetation maps. 7. Integrate: Combine with other GIS data for insights. 8. Share: Export maps for reporting or further study. This concise workflow uses remote sensing and GIS for efficient vegetation mapping.

Data Collection: Gather satellite imagery or aerial photos. Processing: Enhance and georeference the imagery. Classification: Use algorithms to categorize land cover types. Validation: Verify accuracy with ground truth data. Refinement: Adjust results as needed. Visualization: Display classified map in GIS for analysis.

Calculating vegetation indices is an important task in detecting vegetation health, and for this we need GIS or remote sensing softwares to calculate indices like NDVI, SAVI and EVI , we can also use the raster calculator for it by applying the index equation in our satellite bands layers

for vegetation mapping 1st you have to use satellite imagery, which can capture the reflectance of different wavelengths of light from the earth's surface.then calculate NDVI or do classification by which you will get different types of vegetation like forest, grass land, barren land, trees etc after that make a layout

After getting the image of the interested area then, we have to calculate NDVI (Normalized Different Vegetation Index)to correct the image.

Nowadays, many classifiers can be used, however, the input data will always be more important than the used classifier. So, the user needs to do some previous analysis about the study area and estimate and evaluate the best attributes to use. Try different spectral indexes and multi source data is also an option. SAR data can provide information about vegetation water content and structure, while optical data can provide biophysical information. This data combined certainly will bring best results.

One important aspect to consider is the spatial distribution of the validation dataset when performing supervised classification. To improve the robustness of the classified model, I would recommend to choose validation samples from different regions rather than restricting it to one area.

The third step is classifying vegetation type using supervised or unsupervised classification methods by assigning pixels to classes

By classification we can map vegetation there are many applications we can work on this. We can use Ariel Photogrammetry, Lidar technology.

It is possible to use classifications methods (supervised or unsupervised) using GIS or remote sensing softwares (ArcGIS, ENVI), the most important part in supervised classification is to choose the adequate sample to not distort the results

I find that it is important to think about what you're trying to show in the vegetation map and to have a good sense of the study area. For the NDVI map for mangroves in Singapore for my honors thesis, the map of the island wasn't very useful in showing the data as the mangroves were mainly along the coastline, so I had to use a few zoom insets to show the results clearly for comparison between the different areas. A table / bar chart of the mean NDVI values makes it easier for quantitative comparison between the different areas too.

Afin que la carte soit lisible et intuitive, il est conseillé d'utiliser des gradients de couleur qui représentent également des gradients écologiques : des tons rouges pour les végétations les plus sèches vers des tons bleus pour les végétations les plus humides, ou qui représentent la physionomie de la végétation : du plus clair pour les végétations basses, herbacées à des tons plus sombres pour les végétations arbustives à arborées.

It is important to know at what scale (e.g. 10K or 50K or 250K) you want to achieve the results. There are multiple data sources starting from course to fine resolution, Panchromatic, multi-spectral to hyperspectral etc. Hence, source of the data will be varying based on objective as well. In general mapping only vegetation is little easier unlike classifying vegetation. So, one who wants to extract only vegetation can use multispectral images and utilize infra red bands. There are many popular indices like NDVI, SAVI or LAI etc. would do the job. Along with that feature segmentation would help to create an extent of veg cover in the form of geometry which can be used further for GIS analysis and data visualization.

Sometimes less is more. When you have a categorization that you want to impose on a map, the more categories you have, the harder the map is to use. Focus on what you want the visual to highlight and emphasize that.

First it must be evaluate how big is the area size. If the area is small, then we can use phogrammetry to map this area by using UAV(drone)photogrammetry we can get god result. This method is more accurate.

In this type of work, it's important to consider what we want to visualize, with what precision, and what spatial, spectral, and temporal resolution we have in our images. It's pointless to have points measured in centimeters or inches and images in meters or feet!