| ArcGIS REST Services Directory | Login |
| Home > services > DISASTERS_202501_FIRE_CA > 2501_aviris_pca (ImageServer) | API Reference |
| JSON | SOAP | WMS |
Date of Image(s):
1/11/2025
Date of Next Image:
1/16/2025
Summary:
This product visualizes three dominant principal components derived from the surface reflectance estimates from airborne imaging spectroscopy measurements taken by NASA JPL’s AVIRIS-3. The image visualizes the relative strength of three dominant eigenvectors, estimated from data across the full scene. Eigenvectors from this high (284) dimensional dataset do not specifically identify physical features but are correlated to physical phenomena; similar to how ‘red’, ‘green’, or ‘blue’ in a traditional image are correlated to, but not directly indicative of, specific physical processes. Instead of just “red,” “green,” and “blue,” this analysis considers 284 unique “colors” in the visible to shortwave infrared spectrum, providing a much more robust depiction of spectral differences. Thus, contrast between different colored regions in this image can be used to infer strong differences in surface type. For example, the fire boundary at the time of image collection is readily apparent. Explained another way, within a given scene, burned structures will share a unique spectral reflectance signature in “284-color-space” compared to other surfaces and display as the same color in the PCA image.
The product ground spatial resolution is approximately 2.8 m, and flights occurred between 19:40 and 21:00 UTC on Jan 11th 2025. Raw data are available for download here: https://popo.jpl.nasa.gov/pub/LA_Fires/dist/eaton_pca_20250111.tif, https://popo.jpl.nasa.gov/pub/LA_Fires/dist/palisades_pca_20250111.tif.
In the absence of a supervised image classification, the PCA image can demonstrate areas with similar spectral reflectance properties. On visual inspection of the PCA image, we can intuitively differentiate burned areas from non-burned areas, even for different types of burned surfaces (e.g. burned vegetation vs. burned structures).
Note: PCA images are produced on a scene-specific basis. This means unique PCA color representation in one scene does not translate to the same colors in another scene. For example, colors in the Eaton Fire scene do not relate to the same colors in the Palisades fire scene. <o:p></o:p>
Suggested Use:
As an unsupervised method to reduce dimensionality and improve feature extraction in high-spectral resolution datasets, the PCA image can serve as an intermediate visual guide to reveal the information content and diversity of surfaces in the AVIRIS-3 datasets. Analysts may use this image to visually understand the spectral diversity of fire impacted areas, especially in conjunction with ancillary data that characterizes the unique underlying physical phenomena represented by unique colors in the PCA image. <o:p></o:p>
POC:
Philip G. Brodrick (JPL) and David R. Thompson (JPL)
Citation:
AVIRIS-3 Radiance Data:
Eckert, R., D.R. Thompson, A.M. Chlus, J.W. Chapman, M. Eastwood, M. Bernas, S. Geier, M. Helmlinger, D. Keymeulen, E. Liggett, S. Nadgauda, L.M. Rios, L.A. Shaw, W. Olson-Duvall, P.G. Brodrick, and R.O. Green. 2024. AVIRIS-3 L1B Calibrated Radiance, Facility Instrument Collection. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/2356
AVIRIS-3 Reflectance Data:
Brodrick, P.G., A.M. Chlus, U.N. Bohn, E. Greenberg, J. Montgomery, J.W. Chapman, M. Eastwood, S.R. Lundeen, R. Eckert, W. Olson-Duvall, D.R. Thompson, and R.O. Green. 2025. AVIRIS-3 L2A Orthocorrected Surface Reflectance, Facility Instrument Collection. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/2357
Service URL:
WMS Endpoint:
Date of Image(s):
1/11/2025
Date of Next Image:
1/16/2025
Summary:
This product visualizes three dominant principal components derived from the surface reflectance estimates from airborne imaging spectroscopy measurements taken by NASA JPL’s AVIRIS-3. The image visualizes the relative strength of three dominant eigenvectors, estimated from data across the full scene. Eigenvectors from this high (284) dimensional dataset do not specifically identify physical features but are correlated to physical phenomena; similar to how ‘red’, ‘green’, or ‘blue’ in a traditional image are correlated to, but not directly indicative of, specific physical processes. Instead of just “red,” “green,” and “blue,” this analysis considers 284 unique “colors” in the visible to shortwave infrared spectrum, providing a much more robust depiction of spectral differences. Thus, contrast between different colored regions in this image can be used to infer strong differences in surface type. For example, the fire boundary at the time of image collection is readily apparent. Explained another way, within a given scene, burned structures will share a unique spectral reflectance signature in “284-color-space” compared to other surfaces and display as the same color in the PCA image.
The product ground spatial resolution is approximately 2.8 m, and flights occurred between 19:40 and 21:00 UTC on Jan 11th 2025. Raw data are available for download here: https://popo.jpl.nasa.gov/pub/LA_Fires/dist/eaton_pca_20250111.tif, https://popo.jpl.nasa.gov/pub/LA_Fires/dist/palisades_pca_20250111.tif.
In the absence of a supervised image classification, the PCA image can demonstrate areas with similar spectral reflectance properties. On visual inspection of the PCA image, we can intuitively differentiate burned areas from non-burned areas, even for different types of burned surfaces (e.g. burned vegetation vs. burned structures).
Note: PCA images are produced on a scene-specific basis. This means unique PCA color representation in one scene does not translate to the same colors in another scene. For example, colors in the Eaton Fire scene do not relate to the same colors in the Palisades fire scene. <o:p></o:p>
Suggested Use:
As an unsupervised method to reduce dimensionality and improve feature extraction in high-spectral resolution datasets, the PCA image can serve as an intermediate visual guide to reveal the information content and diversity of surfaces in the AVIRIS-3 datasets. Analysts may use this image to visually understand the spectral diversity of fire impacted areas, especially in conjunction with ancillary data that characterizes the unique underlying physical phenomena represented by unique colors in the PCA image. <o:p></o:p>
POC:
Philip G. Brodrick (JPL) and David R. Thompson (JPL)
Citation:
AVIRIS-3 Radiance Data:
Eckert, R., D.R. Thompson, A.M. Chlus, J.W. Chapman, M. Eastwood, M. Bernas, S. Geier, M. Helmlinger, D. Keymeulen, E. Liggett, S. Nadgauda, L.M. Rios, L.A. Shaw, W. Olson-Duvall, P.G. Brodrick, and R.O. Green. 2024. AVIRIS-3 L1B Calibrated Radiance, Facility Instrument Collection. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/2356
AVIRIS-3 Reflectance Data:
Brodrick, P.G., A.M. Chlus, U.N. Bohn, E. Greenberg, J. Montgomery, J.W. Chapman, M. Eastwood, S.R. Lundeen, R. Eckert, W. Olson-Duvall, D.R. Thompson, and R.O. Green. 2025. AVIRIS-3 L2A Orthocorrected Surface Reflectance, Facility Instrument Collection. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/2357
Service URL:
WMS Endpoint: