How to Generate Precise DEMs from SAR Data with InSAR in ENVI SARscape
HTML-код
- Опубликовано: 6 фев 2025
- For today's video on my RUclips channel, which focuses on Space Remote Sensing and GIS, we'll dive into the fascinating process of DEM creation from SAR images using Interferometry with the help of ENVI SARscape. This method employs InSAR (Interferometric Synthetic Aperture Radar) and DInSAR (Differential InSAR) techniques to generate accurate Digital Elevation Models (DEM), as well as Coherence and Land Displacement maps.
Here’s a breakdown of the key steps involved in this process:
1. Co-registration using DEM
The first step ensures the alignment of corresponding points in both SAR images. By using an existing DEM, the processing parameters are fine-tuned to ensure proper image registration.
2. Common Doppler Bandwidth Filtering
If there are discrepancies in the Doppler centroids of the images, this filtering step minimizes de-correlation between them, improving the quality of the interferogram.
3. Interferogram Generation
A spectral shift filter is applied to the image pair. The Hermitian product is then calculated to generate the interferogram, which represents the phase difference between the SAR images.
4. DEM Flattening
Synthetic fringes are generated from a lower-resolution DEM or ellipsoidal height using a backward geocoding approach. These fringes are cross-multiplied by the SAR interferogram to remove low-frequency phase components, setting the stage for phase unwrapping.
5. Adaptive Filtering
The interferogram is filtered to enhance phase signal-to-noise ratio (SNR), ensuring more accurate height data in the final DEM.
6. Phase Unwrapping
The phase of the interferogram is unwrapped using a region-growing approach, converting the wrapped phase data into continuous elevation information.
7. Phase Editing
If there are errors in the unwrapping process, they can be manually or semi-automatically corrected to ensure accuracy in the final DEM.
8. Geometry Optimization with Ground Control Points (GCPs)
The SAR orbit parameters are refined using GCPs to achieve precise phase-to-height conversion. Both manual and automatic GCP identification methods are available for accuracy.
9. Phase-to-Map Conversion
The final DEM is obtained by rigorously converting the phase data into a map, taking into account the range-Doppler nature of SAR acquisitions. This conversion uses all necessary geodetic and cartographic transformations, ensuring the DEM is presented in the correct reference system.
10. Final DEM Creation
The final result is an accurate Digital Elevation Model (DEM), derived from the SAR data, providing detailed topographic information about the surveyed area.
In this video, we will also focus on the importance of using DORIS precise orbits for higher DEM accuracy, rather than relying solely on the orbital parameters provided with the raw data.
Make sure to stay tuned for a step-by-step demonstration using ENVI SARscape to apply these techniques and generate a high-resolution DEM!
Hlo sir if i am doing DinSAR processing in SarScape........and export image in tiff format for map composistion.................
My values for displacement are -0.14 to +0.11 ........ What are its units......if i am using 12 day data.........than this values are for 12 days or year..............
But firstly plz tell me the units of my output.....is it mtrs,cm or mm
Displacement (symbolized d or s ), also called length or distance, is a one-dimensional quantity representing the separation between two defined points. The standard unit of displacement in the International System of Units ( SI ) is the meter (m).
In SARscape processing, specifically for Differential InSAR (DinSAR), the displacement values represent the amount of surface movement detected between two radar acquisitions. These values are typically in meters (m) and reflect the displacement between two radar images, which can be caused by phenomena such as land subsidence, infrastructure deformation, or tectonic shifts.
Units of Displacement Values
The displacement values you have (-0.14 m to +0.11 m) are in meters (m).
This means the surface has shifted by a range of 0.14 meters to 0.11 meters.
Displacement Over Time (12-Day Interval)
Since you are working with 12-day data, these values represent the displacement over a 12-day period. The values are calculated based on the difference between two SAR acquisitions that were taken 12 days apart.
Therefore:
The displacement of -0.14 m to +0.11 m is for the 12-day period.
If you wanted to scale this to a per-year basis, you would multiply by the number of 12-day intervals in a year.
There are approximately 30.42 days per month on average, so a year would have about 365.25 days.
Dividing 365.25 days by 12 gives approximately 30.44 intervals per year.
To convert your 12-day displacement values to a per-year basis, you would multiply the displacement range by 30.44.
For example:
-0.14 m per 12 days × 30.44 = -4.26 m per year
+0.11 m per 12 days × 30.44 = +3.34 m per year
Thus, the displacement over the entire year would be in the range of -4.26 m to +3.34 m, if you scale it up.
Exporting to TIFF
When you export the displacement map to TIFF format for map composition, ensure that you specify the coordinate reference system (CRS) and any relevant metadata that can clarify the time frame and spatial scale for your map's users.
If you need the displacement data to be shown over a specific spatial extent, be sure to handle the geographic projection and scaling appropriately during export.
@spaceremotesensingandgis5626 sir the 4mtr values are very high......for land displacement........values for dispalcement is in. mm .....
If have have to find velocity than i have to divide displacement by time period .....for ex if my displacemnt for 12 days is 0.48mts. Tha to find velocity to do this 0.48/12. Which is 0.4mts. Which is very less value....in comparison to velocity......if we consider 0.48mtr for 1 day than its is suitable for glacier
@@Dr.MeMer5 Its difficult to tell that as I am not much work on glacier. I will check and let you know if I get any information.