Dasymetric Mapping of Population Using Land Cover Data in JBNERR, Puerto Rico during 1990–2010

Round 1

Reviewer 1 Report

This article appears to be well organized and structured and uses appropriate methodologies, however, the referenced literature must be entered with a serial number referred to in the text. The pictures have a serial number with a description behind the picture. The tables have a serial number and a description in front of the figure. Formulas have only a number in parentheses without a description. In lines 317, 321, 386, and 388 an error occurred in the text. In the conclusion, line 477 states 30m2, and it is probably 900m2 or 30m x 30m. Also, in table two add that the values refer to the pixel size (900m2) In the conclusion in line 475 where the key result of this research is defined, there is no need to refer to the literature because there is confusion about the authorship of the key result of the paper.

Author Response

Please see the attachment. 

Author Response File: Author Response.docx

Reviewer 2 Report

The paper describes an attempt to create a raster population map using dasymetric modeling for the area in Puerto Rico from 1990 to 2010. Authors use an IDM approach implemented in ArcGIS script to perform a calculation. 

There are many aspects of the paper that need to be explained, the most important are: 

1. To create multi-year comparable population grids the ancillary data needs to be comparable between years themselves. Based on this paper I do not think that this criterion is fulfilled. Fulfilling this criterion is crucial for creating multi-year comparable maps. 

2. The resultant high-resolution maps should be available to download. There are only shown in a few low-quality figures. 

3. Authors use the centroid method to establish a "typical" population density for each ancillary class. This is a very inaccurate approach. 

4. Authors disaggregate census block population data to the grid at the resolution of 30m. How the error is calculated? To assess the error, disaggregated and original data are required. Usually, for that purpose, the disaggregation from larger units (i.e., block group) to smaller units (census blocks) is performed, and based on such data the error are calculated. 

Other comments: 

In Section 4 many references are missing (Error! Reference source not found). 

I do not think that section 2.2.2 should be part of the paper. 

In its current form, is more like a description of the exercise, than a research paper.  In its current form, I do not recommend it for publication. Before the publication authors have to clearly describe how the error is calculated. Also, the problem of non-comparable ancillary data has to be addressed. 

Author Response

Please see the attachment. 

Author Response File: Author Response.docx