Tc minor fixes #475
Tc minor fixes #475
Conversation
Preview is ready!
You can also view the deploy log or read the documentation. |
| The Tasseled Cap (Kauth–Thomas) transform translates the six spectral bands of Landsat into a three indexes describing greenness, wetness and brightness. These indexes can be used to help understand complex ecosystems, such as wetlands or groundwater dependent ecosystems. The Tasseled Cap Percentiles capture how the greenness, wetness and brightness of the landscape behaves over time. | ||
| The Tasseled Cap transformation used to generate this product implements the coefficients described by Crist et al. (1985), adapted for Landsat surface reflectance data. The Tasseled Cap Transformation (TCT) is a linear transformation that projects multispectral data into a new coordinate system defined by three components: Brightness, Greenness, and Wetness. | ||
|
|
||
| In other words, the transformation reduces a larger number of spectral bands into three composite indices that are easier to interpret and apply in analyses. These indices are designed to capture the dominant modes avenues spectral variation in terrestrial landscapes: |
There was a problem hiding this comment.
@jennaguffogg Is this a typo? "These indices are designed to capture the dominant modes avenues spectral variation in terrestrial landscapes"
| @@ -1,8 +1,13 @@ | |||
| ## About | |||
There was a problem hiding this comment.
@jennaguffogg I'm just wondering, did you talk to Cedric or anyone before rewriting the About section? I think the About section is meant to be approved by a director, and usually doesn't change until a new version is released. That's because the About section is the 'official' description of the product.
There was a problem hiding this comment.
Aaah I didn't know that. I'll ask Cedric about it today, and if it needs approval I'll revert it to the old text until we get that.
| ## Background | ||
|
|
||
| The Tasseled Cap transformation (Kauth–Thomas) is a method used to simplify satellite imagery by converting raw spectral data into three key indices: | ||
| The Tasseled Cap transformation (Kauth and Thomas, 1976) is a method used to simplify satellite imagery by converting raw spectral data into three key indices: |
There was a problem hiding this comment.
I am wondering if here we should use both the reference to Kauth and Thomas 1976 and to Crist, E. P. (1985). If someone stops reading at the first couple of lines, they might think that the DEA product is derived from methods in Kauth and Thomas 1976, but actually below it is specified we used Crist 1985. So maybe keeping both can help avoiding misunderstanding
|
also tcp has the same sensor selection as fcp, could you add it? |
|
|
||
| - Brightness is a weighted sum of all bands and reflects overall surface albedo. | ||
| - Greenness contrasts the visible and near-infrared bands to highlight photosynthetically active vegetation. | ||
| - Wetness contrasts shortwave infrared bands with visible and near-infrared bands to detect moisture in soil and vegetation. |
There was a problem hiding this comment.
all of them are weighted sum of all bands in terms of technique. the text here is suggesting something different essentially among all indices. could you clarify how the interpretation favours specific bands within the context of weighted sum?
| These percentiles summarise the distribution of index values over the calendar year and characterise the lower-bound, central tendency and upper-bound of the measures of Brightness, Greenness and Wetness. | ||
|
|
||
| The Tasseled Cap Percentiles are intended to complement the DEA Water Observations (Water Observations from Space) and Fractional Cover algorithms. DEA WO is designed to discriminate open water, but the Tasseled Cap wetness index identifies areas of water and areas where water and vegetation are mixed together; i.e. mangroves and palustrine wetlands. Similarly Fractional Cover describes proportions of green, brown and bare areas in the landscape, and hence the Fractional Cover percentiles can be used in complement to the Tasseled Cap percentiles. | ||
| Percentiles are used instead of minimum, maximum, or mean values because they are less sensitive to outliers and better suited to non-normally distributed data. This makes them more reliable for detecting subtle or seasonal changes in complex environments. |
There was a problem hiding this comment.
this section is to describe how tcp are generated, could you move it to processing steps section to keep consistency among all products?
| The 10th, 50th, and 90th percentiles are then computed from the full time series of index values for each pixel. These percentiles are statistical summaries, not direct satellite observations. They represent synthetic values that describe the distribution of environmental conditions over the year. | ||
|
|
||
| The resulting dataset is a synthetic summary of observed conditions, designed to support environmental monitoring, classification, and change detection across a wide range of Australian ecosystems. | ||
|
|
There was a problem hiding this comment.
could you keep the lineage section to be concisely on the source datasets and their bands? also move how they're computed to processing steps together with the section Annual summarisation using percentiles?
|
Just checking, is this good to be merged? @jennaguffogg @benji-glitsos-ga |
4 participants
Updates to the TC page to reduce duplicate information and increase readability of overview and description pages.