Estimating the Optimal Threshold for Accuracy Assessment of the Global Ecosystem Dynamics Investigation (GEDI) Data in a Gentle Relief Urban Area
, Renato Fontes Guimarães 1,*, Osmar AbÃlio de Carvalho Júnior 1, Roberto Arnaldo Trancoso Gomes 1, Osmar Luiz Ferreira de Carvalho 2 and Thyego Pery Monteiro de Lima 3
Round 1
Reviewer 1 Report
This reviewer thinks that this paper has adequately corrected.
There are no further revision requests in my view.
Author Response
Dear reviewer,
Thank you for your contribution to our manuscript.
Kind regards,
Reviewer 2 Report
Response to Author Comments:
Thank you for your explanations. I think I got the point of this study now. Let me put it in my words: You want to get the best possible accuracy from the data for – and this is important – accuracy assessment! As this method only works with reference data.
Besides accuracy assessment often suffers from underestimates the accuracy of DEM differences and present not the accuracy of the optimal case. But in general, it is not allowed to reduce measurements until it fits to a normal distribution. But to have both, a good filtered data set AND your RMSE after this KS-normal fit would help to describe the accuracy of height data sets.
Nevertheless, I strongly recommend not to use the word “outlier”. For your 100m – threshold – okay these are outliers. But if you narrow down your data set to +/-3m this is no outlier (could be changes, uncertainties, …) these are measurements!
Attached some discussions and details.
Comments for author File: 
Comments.pdf
Author Response
Dear reviewer,
First of all, We would like to thank all of the suggestions and concerns that certainly contribute to improving this manuscript. Following we show one by one the last requests in the attached pdf file.
Kind regards,
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear Authors,
I'm concerned about your assumption about the probability distribution of the terrain elevation differences. You are assuming the normal distribution, but some authors demonstrated that this process follows the Laplace distribution. So, I'd like you to consider revising your assumption. To find references to the Laplace pdf and DEM differences, search the Internet for "elevation bias of SRTM C-and X-Band Digital Elevation Models".
Author Response
Dear reviewer,
First of all, We would like to thank all of the suggestions and concerns that certainly contribute to improving this manuscript. Following we show one by one the last requests in the attached pdf file.
Kind regards,
Author Response File: Author Response.pdf
