The forum is locked.
The Ocean Color Forum has transitioned over to the Earthdata Forum (https://forum.earthdata.nasa.gov/). The information existing below will be retained for historical reference. Please sign into the Earthdata Forum for active user support.
Topic Special Topics / Inherent Optical Properties Workshop / workshop goals, metrics, homework (locked)
Hello All,
Hope everyone is well. Glad to see some traffic on the Forum.
Thought it'd be timely to review our IOP Workshop goals and roles and pose a question or two to the group prior to heading into the weekend.
GOAL: We would like to achieve community consensus on an effective, *consolidated* algorithmic approach for producing *global* scale, remotely sensed IOP data products.
Within the context of the Workshop, we have the following objectives:
(1) Develop a strategy to retrieve IOPs.
(2) Generate an uncertainty budget for the approach.
The "questions to be answered" that we've posed in the past generally map into these two objectives. If you're tired of hearing about them, just skip to the next paragraph ... For (1), we're interested in: (a) the sensitivities of the inversion method and "constants"/shape vectors; (b) identification of failure conditions within the remote sensing paradigm; (c) strategy for generating Level-3 products (e.g., bin IOPs calculated at Level-2 or calculate IOPs using Level-3 Rrs). For (2), we're interested in: (a) estimation of uncertainties for each product; (b) wavelength requirements (e.g., sensitivity to losing or gaining wavelengths); and (c) definition of valid ranges or appropriate trophic levels for each product.
With regards to your roles, the more you're willing to address, the better. We'd greatly appreciate an engaged group. We plan on running and posting analyses for the next two months. We certainly encourage you all to do the same, for as many or as few topics as suit your interests.
With regards to evaluation data sets, we have the following:
* NOMAD v2 (in situ Rrs and in situ IOPs)
* satellite-to-in situ match-ups (satellite Rrs and in situ IOPs)
* Level-2 satellite scenes
* Level-3 satellite scenes
All likely need consideration in some capacity (and all have their own issues -- e.g., the in situ data are admittedly far from perfect).
For those who've read this far, I have "homework" for the group. These are questions we'd like discussed well before we gather in Oct.
(1) What products to consider? a, adg, aph, bb?
(2) What metrics will we adopt to evaluate performance?
Considerations (and possible options) for metrics:
* What are valid ranges for each IOP product?
* Require all retrieved IOPs (adg, aph, bbp) to be valid for a given pixel?
* Should Rrs be "screened"/evaluated prior to IOP calculation?
* Comparisons of spatial coverage?
* Apply homogeneity tests?
* Comparison of time-series and frequency distributions?
Early next week, Sean will be posting our first draft of satellite-to-in situ match-ups (for most of the algorithms). We will also propose a brief time-line, describing what we (OBPG) hope to accomplish analysis-wise in the next two months.
Take care.
Hope everyone is well. Glad to see some traffic on the Forum.
Thought it'd be timely to review our IOP Workshop goals and roles and pose a question or two to the group prior to heading into the weekend.
GOAL: We would like to achieve community consensus on an effective, *consolidated* algorithmic approach for producing *global* scale, remotely sensed IOP data products.
Within the context of the Workshop, we have the following objectives:
(1) Develop a strategy to retrieve IOPs.
(2) Generate an uncertainty budget for the approach.
The "questions to be answered" that we've posed in the past generally map into these two objectives. If you're tired of hearing about them, just skip to the next paragraph ... For (1), we're interested in: (a) the sensitivities of the inversion method and "constants"/shape vectors; (b) identification of failure conditions within the remote sensing paradigm; (c) strategy for generating Level-3 products (e.g., bin IOPs calculated at Level-2 or calculate IOPs using Level-3 Rrs). For (2), we're interested in: (a) estimation of uncertainties for each product; (b) wavelength requirements (e.g., sensitivity to losing or gaining wavelengths); and (c) definition of valid ranges or appropriate trophic levels for each product.
With regards to your roles, the more you're willing to address, the better. We'd greatly appreciate an engaged group. We plan on running and posting analyses for the next two months. We certainly encourage you all to do the same, for as many or as few topics as suit your interests.
With regards to evaluation data sets, we have the following:
* NOMAD v2 (in situ Rrs and in situ IOPs)
* satellite-to-in situ match-ups (satellite Rrs and in situ IOPs)
* Level-2 satellite scenes
* Level-3 satellite scenes
All likely need consideration in some capacity (and all have their own issues -- e.g., the in situ data are admittedly far from perfect).
For those who've read this far, I have "homework" for the group. These are questions we'd like discussed well before we gather in Oct.
(1) What products to consider? a, adg, aph, bb?
(2) What metrics will we adopt to evaluate performance?
Considerations (and possible options) for metrics:
* What are valid ranges for each IOP product?
* Require all retrieved IOPs (adg, aph, bbp) to be valid for a given pixel?
* Should Rrs be "screened"/evaluated prior to IOP calculation?
* Comparisons of spatial coverage?
* Apply homogeneity tests?
* Comparison of time-series and frequency distributions?
Early next week, Sean will be posting our first draft of satellite-to-in situ match-ups (for most of the algorithms). We will also propose a brief time-line, describing what we (OBPG) hope to accomplish analysis-wise in the next two months.
Take care.
Topic Special Topics / Inherent Optical Properties Workshop / workshop goals, metrics, homework (locked)