Today we will be processing some Radar data using a program called SNAP, an OpenSource application released by the European Space Agency for processing Sentinel data, and can perform most of the remote sensing tasks we have done to date.
Start by opening snap and we will take a look around at some of the features
The general layout of the program is as follows
Toolbars along the top are generally categorized by data type, as such today we will focus primarily on the Radar options.
Along the left, the top panel is “Product Explorer” and is the listing of all of our open images, when including all the metadata we will refer to these images as “Products” one cool feature of SNAP is that we do not need to extract ZIP files of products that we download and can use them as an individual file. The bottom panel has several tabs and will be used primarily for navigation, as well as displaying the results of the analysis.
You will also want to take note of the tabs along the right edge of the window, specifically the Product Library tab, this is the way we will access our data. Another handy feature of SNAP is the ability to download data directly inside the application. To enable this option if you wish go to Tools > Options > General > Product Library, and you can add your accounts to online data repositories (Note that USGS requires requesting API access before it will function). However for today, we have some already downloaded data in a “Local Repository”.
You can find the data for today’s lab by opening Product Library, Choosing “All Local Folders” as the repository, and then using the + to add a new local folder. The folder you will be adding is L:\DATA\SNAP, when it asks do not generate quick views or search recursively.
You will now be able to search this folder and get a list of all locally available data, and the footprints will be visible in the Area of Interest, highlighted in red when selected.
This search can be narrowed in a few ways, Mission Allows us to select which satellite we are interested in. We can also specify dates of capture. and Right-clicking on the Area of interest will allow you to draw a polygon to define the area to search.
For today’s Lab find the Sentinel 1B data over Vancouver Island capture dates are March 19, 2020, and March 31, 2020.
You can open the products by right-clicking and choosing open.
Looking at the data
Start by opening the aquisition from March 19th, expanding to bands, then either double click Amplitude VH or Right-click and Open Image Window, next do the same for Intensity_VH. Now that both bands are open you can click and drag one of the tabs to the side and it will let you split the view.
We also have the option of showing 3 bands as an RGB view. You will find this option under “Window” and “Open RGB Image Window”, note unlike PCI we do not have an RGB mapper so choosing a different selection of bands requires creating a new view.
As with the optical imagery we have previously used we need to apply some corrections to our data before moving forward. The process of applying corrections to the Radar data is extremely easy thanks to SNAP being able to use the downloaded metadata, as well as downloading information on the fly from ESA’s servers to set most of the parameters.
The first step is to apply the orbit file (in the Radar menu), making sure to change the directory to your K: drive. By default the Name will be the same as the original file with the addition of the _Orb suffix, however, you could change this to a friendlier name if you wish.
After the orbit file is applied we will perform radiometric calibration, on our new files, this option can be found in Radar > Radiometric > Calibration. Again make sure to save to your K drive.
Finally we will coregister the data. at this stage is it important to watch our order of operations as we want the older data to be our Master for coregistration. Open the Coregistration tool from Radar > Coregistration > Coregistration. Start by dragging the March 19th scene into the product selection, followed by March 31st, your window should appear as below.
By this stage you may have noticed that the image is flipped, to fix this we will go into Geometric Corrections, and Range Doppler Terrain Correction, go into the Processing Parameters and ensure that ‘Mask out areas without elevation’ is not checked. We want to wait as long into the process as possible to apply the geometric corrections, as resampling the bands can cause some loss in detail.
In the radar menu, there is a section for SAR Applications, Let’s try a couple of these out, as these take a relatively long time to compute we will create a subset of our data, under the raster menu you will find the subset command, create a subset including Vancouver similar in extent to the image below. Create the subset from one of your _orb files, that is a file before radiometric calibration has been applied.
under the Ocean submenu, you will find Ocean Object Detection. Use this tool on one of your new subsetof the ocean surrounding Vancouver.
Once completed if you zoom in you’re should be able to see circles around the ships in the ocean.
This week’s lab assignment is simple, export the RGB view of Radiar as an image.