A Simple Procedure to Preprocess and Ingest Level-2 Ocean Color Data into Google Earth Engine
Round 1
Reviewer 1 Report
1st comment: all along the text, the authors refer to the OC data they use as “High resolution OC data”: this is misleading as such data are considered by the community of remote sensing scientists, including the OC scientists, as “Medium Resolution”; it is obvious that the current OC sensors such as MODIS, VIRS, S3/OLCI, … deliver products at higher resolution than SeaWiFS (at 1 and 4 km resolution), but they do not qualify for the HR designation which is only valid for Landsat or Sentinel-2 (5-to-30m, in waiting for VHR data at m scale that could be used for OC). We suggest the authors to review the HR qualification.
2nd comment: the scheme, i.e. “swath reprojection”, transforms Level-2 products, according to the NASA classification of EO-products, in Level-3 products; the authors refer to Level-2 (L2) for the original data sets but not to Level-3 for the output, except in §4 Discussions. We suggest adjusting the rest of the text accordingly.
3rd comment: the § on the data ingestion flow describes the work in details; because it is more about the ‘engineering’ in a GIS, we would suggest putting all the details in an annex and keeping the core text more general.
4th comment: the metrics used to compare the original L2 products and the re-gridded/remapped products (L3) are only statistics on pixel-values and smoothness of the fields; this may not be enough for OC scientists and people using the results for BGC modelling (the authors have referred to one paper by Scott & Werdell, 2019, RD-25) and it would be interesting to link the metrics to specific applications which are not those related to scientific research.
5th comment: there are several qualitative statements in the discussion, not really supported by the results. The statement that level 3 OC products are at 4,64 km resolution is misleading : OC level 3 products may exist at different resolution. There is no specific gap to fill. It would be more correct to simply show that GEE is an interesting tool among various other GIS tools. There is no demonstration of benefit for regional algorithm development or OC validation : "immensely benefit" is only a statement.
Author Response
Thank you for the comments. We appreciate the time to review our manuscript and provide comments for improvement.
We provide the response to reviewer's comments in the PDF. Please see the attached file for details.
Thank you!
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors present an approach for ingesting Ocean Color data into GEE, and they provide code to help researchers perform this task. Statistical comparisons of the reprojected, ingested data with the input level-2 data obtained from the NASA or JAXA portals show that the dataset attributes are approximately preserved by the steps proposed in the manuscript. The manuscript is well-written and the topic should be very useful to the ocean color community, wherein the opportunities provided by GEE have generally been under-utilized.
I have minor comments to the authors, which they may consider, as follows:
1) My understanding is that a 2x reduction in spatial resolution is necessary for MODIS imagery in order to prevent salt-and-pepper-like artifacts (e.g., Fig. 2 Remapped Edge-R panel). If this is correct, can the authors clarify in the abstract that the final output achieved for full-swath width imagery is reduced by half relative to the L2 imagery?
2) Near the top of page 2, the authors mention land-based scientific accomplishments using Landsat in GEE. The same could be said about applying GEE high resolution Landsat and Sentinel-2 archives to study coastal ocean processes, for example kelp forest and salt marsh systems. If they wish, the authors could clarify that high resolution data from Sentinel-2 and Landsat are presently available in GEE at native resolution and are applicable for some coastal ocean research topics.
3) Ocean color data products from legacy (e.g., MODIS) platforms are often unsatisfactory for coastal ocean research due to low spatial (e.g., 1km) resolution or inadequate atmospheric correction (e.g., negligible NIR signals) in optically complex and heterogeneous coastal water bodies. But even for open ocean research, GEE has also not been widely utilized (so far) by the ocean color community, despite the benefits for processing large-scale datasets. The authors could mention these points to add some nuance to their discussion on page 2 of why ocean color practitioners have not used GEE in coastal OC research.
Thanks to the authors for their clear presentation and for releasing their code. I recommend publication of this manuscript.
Author Response
Thank you for taking the time to review our manuscript and for offering constructive comments for improvement. We really appreciate it.
We provide the response to reviewer's comments in the PDF. Please see the attached file for details.
Thank you!
Author Response File: Author Response.pdf
