An Analysis of Urban Land Use/Land Cover Changes in Blantyre City, Southern Malawi (1994–2018)
Round 1
Reviewer 1 Report
The author(s) have addressed most of the problems identified in the previous review. Suggestions from the reviewers have also been taken into consideration. Nevertheless, this manuscript still has the following issues that need to be addressed.
(1) (Line 120) Its topography ranges from an elevation of about 780 to 1,612 m.a.s.l (meters above sea level Please do not use the abbreviation the first time a term is introduced.)
(2) The color ramp of elevation in Figure 1 is still problematic. The light-yellow color (between the cyan color and the light green color) and the yellow color (between the green color and the orange color) do not represent continuous elevation change. It is recommended that the author(s) use continuous color schemes for elevation.
(8) It is unclear how the political transition impacted the LULC change and urban growth in Blantyre. The paper indicates that the forest reserves had been protected during the “one-party state” political regime but experienced destruction and problems in the democratic era. The author(s) need to elaborate on why there had been such changes.
(4) There are typos and grammatic problems in this manuscript. The author(s) need to carefully do another round of proofreading and editing to improve the language. A few examples are listed as follows.
(The year of) Year 2018 was chosen as the end 96 date because it is when the field verification was taken while 2007 was chosen as an intermediate date 97 to illustrate the rates of change. (Line 96).
Reference was also made to the google earth (Google Earth) archived images and the ground control points (Line 159).
Line 193-198 (please pay attention to the revised punctuation): From the transitional matrices, the diagonal values in each matrix indicate persistency between LULC 193 classes from initial time ?1, and the later time ?2, while the off-diagonal entries indicate the 194 transition from one LULC class to the other. The gain was also calculated through the difference 195 between the total value of each LULC class for the later period ?+?, and persistency ???; while the 196 loss was the difference between the total for the initial time ??+, and persistency ???.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Review of “An analysis of Urban Land Use/Land cover changes in Blantyre City, Sothern Malawi (1994-2018)”
This study uses multi-temporal satellite data to quantify changes in land uses in Blantyre City, Malawi. While there is not much new about the methods employed, the data that the paper provides for an important East African city and its land use experience is invaluable. The study is competently done and the paper well written. The addition of an accuracy analysis to the results has strengthened the paper considerably. I believe it is ready for publication, but will list some minor editorial changes that the authors may wish to consider.
Page 1 Abstract line 10: where urbanization -> where the urbanization
Page 1 Abstract line 14: imageries of 1994… -? Imagery from 1994..
Page 1 Abstract line 14: and the Geographic -> and a Geographic
Page 1 Abstract line 21: the case with -> the case with the
Page 1 line 40: in the informal settlement -> in informal settlements
Page 2 line 59: in the informal -> in informal
Page 2 line 67; into uninformed -> in uninformed
Page 2 line 82: infrastructures -> infrastructure
Page 2 line 85: from the unclear -> from unclear
Page 2 line 85: mandate -> mandates
Page 2 line 92: enough of empirical -> enough empirical
Page 2 line 94: city remotely -> city using remotely
Page 6 Equations 2: Why is the second major term a natural logarithm? Isn’t this a simple proportion?
Page 10 Table 5: Appears to have a formatting error
Page 14 Figure 4: In legend: vegetation -> vegetation
Page 15 line 340: caused the decrease in the groundwater -> cause a decrease in groundwater
Page 15 line 342; flood -> flooding
Page 16 line 367; and avoid -> and avoids
Page 16 line 369: is protected -> are protected
Page 16 line 372: vegetation class and avoids -> the vegetation class and avoids the
Page 17 line 405: by increase -> by an increase
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 3 Report
Authors made a major revision of the paper, lot of issues have been revised and corrected. Therefore, the paper significantly increased quality level. I reccomend it for publishing in journal.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
The article describes land cover changes in Malawi during 1994-2018. Generally, paper has an appropriate structure for academic paper and describes an interesting topic for readers.
I am sorry, but according to mi opinion, the article is NOT suitable for publishing in journal with IF 2.5 because of very low academic level. It relies on simple/basic analysis only, the implementation of GIS is at level of bachelor study, there is no novelty, all steps are made by routine in the field of GIS. I feel a major lack of basic knowledge of GIS in this artictle and I would like to reccomend to authors to cooperate with a GIS specialist or cartographer.
Moreover, authors operate inappropriately with "land use" and "land cover". It is not the same, it is not permissible to confuse both names! I strongly reccomend to authors to separate and define what exactly they observed in the article. Finally, it is not appropriate to use abbreviations in abstract (LULC, line 14), without description.
Why authors used ArcGIS v10.2 (while current version is 10.7) ? Finally, I have a few comments about figures/maps:
All figures: If the map is oriented to the north, then north arrow is not necessary.
Figure 1 - why this inappropriate colour scale was used for elevation? there are "standardized" colour scale for elevation. I do not see Blantyre City in the map.
Figure 3 - it seems that these two maps could be used for comparison - if you use the same colour scale, the minimum and the maximum values must be the same
Figure 4 - why the figure is placed here? I expect, that readers of high-level academic journal know how built-up area or water look like
Reviewer 2 Report
This article used the classified maps obtained from two Landsat images (1994 and 2018) to investigate the urban development in Blantyre City, Malawi. According to the comments below, I am afraid to reject this article.
Overall, the authors just conducted a routine image classification using maximum likelihood classification (MLC) and found the rate of urban development in the target area. Accordingly, the utmost issue of this article is the lack of novelty in methodology and implications. The content and contribution of this research article are almost a decade old. We live in the time of Google Earth Engine which by the researches are able to process hundreds of images in a short time, so the classification of just two images is not considered a big deal. While (MLC) is a reliable method, however, it has lower accuracy in comparison to other classification methods such as support vector machine, decision tree classification, and etc. So, Authors are encouraged to use other classification methods in particle for research articles. The authors should provide more details about radiometric and atmospheric corrections. Why did authors first classify vegetation using NDVI and again classify the vegetation using the MLC method? Why did authors select those specific years (1994 and 2018) as a classification period? The number of ground control points are is not provided.
Reviewer 3 Report
This is a straightforward but well executed multitemporal land use change study of Blantyre in Malawi. There is a shortage of good papers on African land use, so this is a welcome addition. However, the authors should make some edits that could further improve the paper.
Figure 6 shows that some urban land in 1194 was not urban in 2018. Were slums moved? Is this an error? See the easter extreme of the study area, and the south eastern extreme.
Section 4.4 would be a good place to summarize the accuracy assessment of the classification that was performed.
My specific suggestions follow:
Page 1 line 37: settlement, higher -> settlements, a higher
Page 2 line 49: country -> countries
Page 2 line 66: expedient -> expedited
Page 2 line 77: often lacks -> often lack
Page 2 line 78:mechanism. ->. mechanisms.
Page 2 line 90: is in three folds -> is threefold
Page 3 line 101/2, four decimal places of seconds implies 3mm accuracy. Round!
Page 3 Figure 1: Use a conventional topography color sequence. Avoid red, which conflicts with the roads.
Page 3 line 116: ragged -> rugged
Page 4 Sec. 2.1.2 Is this climate detail relevant? At the very least it could be reduced to one sentence.
Page 10 line 237: for 1994 -> for th e1994
Page 10 line 248; minimal loss--see comment above
Page 11 line 265: vegetation class -> the vegetation class
Page 11 line 271: inferred -> imply
Page 11 line 273; gain from -> gain more from
Page 12 lines 286, 288, 299, 302, 303: put "the" before each class name
Page 12 line 290; infers -> implies
Page 12 line 302: the nutshell -> a nutshell
Page 14 line 326: avoid -> avoids
Page 14 line 328: Mudi catchment areas -> the Mudi catchment area
Page 14 line 330: built up -> built up land
Page 14 line 333: is existence -> exist
Page 15 line 357: which -> of which
Page 15 line 359: the role to supply land -> the supply of land
Page 15 line 368: as area -> as being
Page 15 line 370: on marginalized -> in marginalized
Reviewer 4 Report
This manuscript analyzes urban land use/land cover (LULC) change in Blantyre city, Malawi. As one of the few studies that focuses on LULC in sub-Saharan African cities, this research is of great importance. Nevertheless, the paper has the following problems that need to be addressed.
(1) Urban LULC changes are rapid and dynamic in developing countries like Malawi. This study covers a time span of 24 years (1994-2018). However, this research utilizes only two remote sensing images in 1994 and 2008, which cannot adequately illustrate the LULC change dynamics in such a long period of time. At least one more image on an intermediate date around the mid-2000s is needed. An additional image might also help to validate the hypothesis of “a two-step temporal change firstly from vegetation to bare land and later from bare land to built-up” (Line 331).
(2) Post-classification change detection should be built upon high accuracy rates of classified images. In this study “Accuracy assessment for 1994 classified map was not conducted due to lack of reference data” (Line 238). Even if “all classification processes were similar to the 2019 LULC classified map” (Line 239), without accuracy assessment no one can guarantee the 1994 classified map has a satisfactory accuracy rate. The validity of the change detection and following LULC analyses is highly questionable.
(3) In Figure 1, the color ramp of elevation is problematic. The authors use a random color ramp for elevation, in which two close colors (bright tangerine and red) actually indicate dramatic different elevations.
(4) Section “2.1.2. Climate” seems unneeded in this manuscript. How is it related to this study? If it is only used to justify the selection of remote sensing images in the dry season, one or two sentences would suffice.
(5) The title of Figure 3 does not really reflect the content to the two maps. Maps a) 1994 and b) 2018 show the changes in NDVI in Blantyre. Without any discussion about the relationship between NDVI and land cover (specifically vegetation), the authors cannot directly use NDVI to indicate vegetation.
(6) The two maps in Figure 5 are just false color composite of remote sensing images. They are not classified maps with different land use/land cover. Consequently, they cannot reflect any LULC change.
(7) P+j , Pjj, and Pii are included in Equations (3) and (4). Nevertheless, the authors do not explain what they are.
(8) The title of this article emphasizes LULC changes during the Democratic Era (1994-2018). However, the authors fail to clearly explain how the political transition “from a one-party state to a democracy” (Line 355) impacted the urban growth, especially in terms of housing construction, in Blantyre.
