Specify projection when exporting images

[zwy1502]

Here is my code, which retrieves the Sentinel 2 image for a specified time period. However, I have found that some of my images are projected as EPSG32636, while others are 32635. I would like to use the following code to export the images:

geemap. download_ee_image collection (s2composite),
out_dir=r"F:\nlhS2data\data1101_1105",
scale=10,
crs='EPSG:32636')

My question is, do I need to specify the CRS parameter now? I'm worried that specifying different projections may cause offset, so I want to export the original Sentinel2 image. I think the method provided by Geemap is very simple. Can Teacher Wu teach me? Should I set this parameter? Is it still not set?

[giswqs]

This is as expected. Sentinel-2 uses UTM projection. The EPSG code changes depending on the location. If you don't specify crs, it will export the original projection of the image.

[zwy1502]

This is as expected. Sentinel-2 uses UTM projection. The EPSG code changes depending on the location. If you don't specify crs, it will export the original projection of the image.

Teacher Wu, I would like to discuss a few more questions with you and look forward to your reply. I have had the opportunity to purchase and read a book about Geemap published by Teacher Wu on the CSDN website. In Chapter 7, "Exporting GEE Data," there is a sentence that says: Within a few seconds, you should see a file called Landsat.tif in the current working directory. If only the scale parameter is specified without specifying the crs and crs_transform parameters, Earth Engine will calculate a crs_transform parameter for you. However, this may cause the image to shift relative to another image with the same pixel size. Therefore, the exported image may not align with the origin of the projection. Let's retrieve the projection information of the image so that we can use it to specify the crs and crd_transform parameters.

My current requirement is also to export images from a collection of images. My problem is that I don't know how to understand what you said, which has caused me to have more concerns about whether to specify CRS. Can you please describe in detail when CRS and crs_transform parameters need to be specified? I remember what you said in class was that pixel alignment is very important when you need to analyze a time series, because not specifying it can lead to a certain offset. What the teacher meant was that when I download images from a certain time period, such as all images from the same region within a year, I should specify the same CRS and crs_transform parameters for these images, right? Otherwise, simply specifying the scale parameter in the same area may not necessarily align pixels, but like me, it only filters a certain type of image within a certain area. Is it not necessary to specify a specific CRS?