Using EO Browser I’ve downloaded the NDVI index,
why the geotiff 32bit has 3bands and not a unique band with the NDVI value of the pixels ?
Is possibile to obtain a one band raster? Or which of the 3 bands is the correct one ?
many thanks
Simone
Using EO Browser I’ve downloaded the NDVI index,
why the geotiff 32bit has 3bands and not a unique band with the NDVI value of the pixels ?
Is possibile to obtain a one band raster? Or which of the 3 bands is the correct one ?
many thanks
Simone
Hi Simone
You get 3 bands because NDVI layer in EO browser is a visualized layer therefore has RGB bands.
Yes it is possible to download only 1 band with actual NDVI values on EO browser.
Please refer to this thread for a detailed answer.
You need to first create your custom layer using this script below which returns 1 NDVI band
//VERSION=3
function setup() {
return{
input: [{
bands: ["B04", "B08"]
}],
output: {
id: "default",
bands:
}
}
}
function evaluatePixel(sample) {
let ndvi = (sample.B08 - sample.B04) / (sample.B08 + sample.B04)
return [ ndvi ]
}
hope that helps, otherwise feel free to ask further questions
Hi @meteoclima,
Sorry, there was a typo in @dorothyrono’s code: it is missing the parameter that sets the output number of bands, which in this case should be 1.
If you try the following Evalscript, it should work (see example):
//VERSION=3
function setup() {
return {
input: ["B08", "B04"],
output: { bands: 1 }
};
}
function evaluatePixel(sample) {
var ndvi = (sample.B08 - sample.B04) / (sample.B08 + sample.B04);
return [ndvi];
}
Note: this script in itself is designed for display purposes (scaled to the range 0-255). However, when you select Image format: TIFF (32-bit float)
in the download section of EO Browser (under Analytical
) the system will retrieve the raw values in Float for you.
Hello,
thank you very much. Yes It works fine! great!
Best Regards
Simone
Thanks
much apreciated
works for normalised burn ratio and helped me understand how the code works ``
//this one is for landsat 4-5 but its always nir-swir/nir+swir eg sentinel2 would be B08-B12/B08-B12...
//VERSION=3
function setup() {
return {
input: ["B04", "B07"],
output: { bands: 1 }
};
}
function evaluatePixel(sample) {
var nbr = (sample.B04 - sample.B07) / (sample.B04 + sample.B07);
return [nbr];
}
//remember when you download the image to select geotiff and 32bit