Radiometric Cross-Calibration of GF-1 PMS Sensor with a New BRDF Model
Round 1
Reviewer 1 Report
1) Results of this study are only based one-time overpass over the two sites and this is simply not enough to draw any conclusion based on such limited data. As the manuscript indicated, the GF-1 satellite was launched on April 26, 2013. Thus, there is a enough long period time of data sets (more than five years) that can be used for the analysis.
2) There is no description of how on-orbit calibration of the PMS sensors is conducted. Therefore, results of trends for the two sites over the entire mission are important to illustrate how stable these sensors are maintained on-orbit.
3) If there is no on-orbit calibration is applied for the PMS sensors, it would be interesting to show how the cross-calibration coefficients change over time. And also, how much improvement it would be for the stability after apply the cross-calibration coefficients to the PMS measurements.
4) What is the official site calibration coefficients? Are they just the coefficients derived previously with similar approach?
5) Do you have prelaunch-based (obtained in lab conditions) calibration coefficients? It would be interesting to use these coefficients and comparison with the results from the cross-calibration coefficients.
Author Response
Manuscript ID: remotesensing-446150
Title: Radiometric Cross-Calibration of GF-1 PMS sensor with a new BRDF model from Hyperion Images
Authors: Qiyue Liu, Tao Yu and Hailiang Gao.
Journal: Remote Sensing
Dear Editor,
We thank the reviewers for the exhaustive review of our paper, constructive comments, corrections and suggestions. A major revision of the paper has been carried out to take all of them into account. Now in the process, we believe the paper has been significantly improved.
All the comments have been addressed in the revised version of the manuscript and highlighted with yellow highlight colour and explained below.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In this work, the authors present their work on the cross-calibration of the GF-1 PMS sensor with the Landsat 8 OLI sensor and Terra MODIS. Previous work had been done to do the cross-calibration between these instruments, but in this work, they introduce a new BRDF model correction which is used to improve the accuracy of the cross calibration results. While previous work used either ground-based measurements or existing BRDF data products, the goal of this work was to develop methodology to construct a BRDF model without the availability of such measurements or data products. To do this, the authors used a kernel-driven BRDF model with inputs from satellite measurements, in this case from Terra MODIS. The BRDF model is then used to correct the reference sensor data, after which it is used to cross-calibrate the PMS sensor. The model was tested over two sites, one with ground-based measurements and the other without these measurements. The BRDF correction showed improvement in the comparisons of the MODIS/OLI data to the PMS data in both cases.
The implementation of the model is straight-forward and shows good results. However, there are several additions/revisions that I would like to see addressed before publication which are listed below.
- There are several typos throughout the manuscript, particularly for several of the acronyms. For instance, there is a typo for MODIS in the last line of the abstract, or RRDF instead of BRDF in the last paragraph of the introduction. Please check the acronym spelling throughout.
- There is also an issue of subject/verb and number agreement in several sentences throught the text which make it difficult to read at times.
- In the introduction (Line 100), you mention some drawbacks to the Terra MODIS BRDF products. Since you are using Terra MODIS to develop your own BRDF model, can you expand on what these drawbacks are and how your new model will be an improvement over using the existing products.
- In Table 1, the solar irradiance of the green channel is too large by an order of magnitude for MODIS.
- Line 180 mentions that ground and atmospheric parameters are used from the Dunhuang site for the Golmud site. However, it appears that only the surface reflectance is used as stated in line 203. The atmospheric parameters were "assumed", but it is not stated how they were chosen, especially since the Golmud data listed in Table 2 is from a different season. It states that a mid-latitude winter model is used, which may explain the differences, but that is unclear. Is atmospheric data from Dunhuang used at all to determine these parameters?
- What sensitivity does the BRDF measurements have to the atmospheric parameters for the VIS and Water Column? Line 215 states that for clear days the atmospheric parameters will not change significantly, but in Table 2, in the 5 days between the MODIS and OLI measurements, the visibility has changed by a factor of 3 and the water column has changed by a factor of 1.6.
- Can you expand/give an example of the Envelope BT calculation? It seems that it is shown in the results of Figure 5 with how the sunny day data in red and is selected to be near the top of all of the data, but given the variation in this data, it is not entirely clear to me how the BT curves which are used to make the EBT are calculated from the raw data.
- I think things would be more clear if you re-organized sections 3 and 4 by discussing the results in section 4.1 right after the description of the BRDF model in section 3.1. Then, the discussion of the gain calculation would lead directly into the results in table 4. In my read through, I had to flip back and forth in order to understand things clearly as the topics changed to previously covered areas.
- In Figure 5, is the temperature unit in degrees C or is it in reference to another temperature value (since it looks close to 0 at t=0, this could be a coincidence though).
- Figures 5 and 6 reference a "Geermu" site instead of Golmud. Please ensure that you have provided the correct results for the manuscript.
- I suggest splitting Table 4 into two tables and adding a more detailed description for each. The first half of the table compares the two sensor cross-calibrations and not to the "official" calibration results, which are quoted in the text and in the caption. Alternatively, add more description to make this distinction.
- Its clear from Table 4 that the correction to the MODIS comparison is larger due to the large view angle of MODIS. It could be interesting to see some kind of correction as a function of view angle plot for the dates of the measurements in questions in order to understand how much impact the BRDF has.
- In Section 5.1, I would like to see more detail on the changes in the parameter values used when estimating the error. The values chosen
seem to be arbitrary. One might use values of associated with the typical variation when doing this type of analysis, and if that is
what you have done, it should be stated.
Author Response
Manuscript ID: remotesensing-446150
Title: Radiometric Cross-Calibration of GF-1 PMS sensor with a new BRDF model from Hyperion Images
Authors: Qiyue Liu, Tao Yu and Hailiang Gao.
Journal: Remote Sensing
Dear Editor,
We thank the reviewers for the exhaustive review of our paper, constructive comments, corrections and suggestions. A major revision of the paper has been carried out to take all of them into account. Now in the process, we believe the paper has been significantly improved.
All the comments have been addressed in the revised version of the manuscript and highlighted with yellow highlight colour and explained below.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I really appreciate your detailed responses and I think your inputs have significantly improved the quality of the manuscript. It would be helpful for user community that what is the impact of the PMS1 gain adjustment after September 2016, in comparison with the period before the gain adjustment.
Author Response
Responses to the Reviewer(s) Comments
Manuscript ID: remotesensing-446150
Title: Radiometric Cross-Calibration of GF-1 PMS sensor with a new BRDF model from Hyperion Images
Authors: Qiyue Liu, Tao Yu and Hailiang Gao.
Journal: Remote Sensing
Dear Editor,
We thank the reviewers for the exhaustive review of our paper, constructive comments, corrections and suggestions. The revision of the paper has been carried out to take all of them into account. Now in the process, we believe the paper has been significantly improved.
All the comments have been addressed in the revised version of the manuscript and highlighted with yellow highlight colour and explained below.
Author Response File: Author Response.pdf
