File talk:Dora 2017-06-26 2020Z.jpg

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Hello Typhoon2013 and MarioProtIV. I have decided to initiate a discussion about our conflict regarding this image. I have already written about this on Typhoon2013's talk page, but in hindsight I should have done that here instead. Anyway, I have a number of points for why I believe the most recent image I uploaded is more suitable for use on the 2017 Pacific hurricane season article than the one that was originally uploaded by Typhoon2013, that has since been backed by MarioProtIV. I do not wish to partake in discussion regarding matters related to user conduct on this talk page (just in case anyone was intending to do so), as this is not the appropriate place. It would not serve any useful purpose in assessing the suitability of one image over the other due to the inevitable creation of a personal and emotionally-charged debate rather than the proper, civil, academic, objective tone which is most desirable and of utmost importance in the current situation. The discussion should be limited solely to the attributes of the images at hand, and evidence which supports each contributor's views.

Firstly, I would like to define specifically which images this discussion relates to. There are three in particular:

  • "The original image":
    • Uploaded by ChocolateTrain as the first image for Hurricane Dora
    • Photo taken by NASA's 'Aqua' satellite at 2020 UTC on 26 June
    • Uploaded at 23:57 UTC on 26 June
  • "The opposing image":
    • Uploaded by Typhoon2013 as a preferred alternative to the original
    • Photo taken by NASA's 'Suomi NPP (SNPP)' satellite at 2100 UTC on 26 June
    • Uploaded at 06:49 UTC on 27 June
  • "The compromise image":
    • Uploaded by ChocolateTrain as an attempt at a (second) compromise
    • Photo taken by NASA's 'Aqua' satellite at 2020 UTC on 26 June
    • Uploaded at 04:57 UTC on 29 June

A good way to determine which image is best is to identify the pros and cons of each 'candidate', and to then evaluate each candidate based on these pros and cons, taking into account the weighting or importance of each point. Below I have drawn up a table listing all of the pros and cons I can think of (and I spent a while thinking of them) regarding each image's attributes. Please do not add directly to the table below if you wish to respond, but rather create your own or just write your reasons in a paragraph.

Image Assessment
Image name Initial uploader Pros Cons
Original ChocolateTrain
  • The cyclone is shown to the viewer without other unrelated cloud formations as distractions
  • All spiral banding features are clearly visible
  • Many pixels over a small geographical area, meaning high resolution
  • Richness of colour and hue
  • The image contains no distinguishable topographical features for the viewer to determine the cyclone's geographic size (too zoomed in)
  • The cyclone eye, though clearly visible, is slightly cloud-filled
Opposing Typhoon2013
  • The image is zoomed out so that Baja California Peninsula can be used as a scale reference with which the viewer can determine the size of the cyclone
  • The cyclone eye is clearly visible and is slightly more distinct
  • Spiral banding features are clearly visible (except for the very right-hand portion)
  • A line runs straight through the image from one-third of the way across the top to the bottom right-hand corner, ruining the picture by fragmentation, and blocking the view of the outer edge of the spiral banding features
  • Unrelated cloud formations are present as a distraction everywhere, particularly in the bottom part of the image, and create a somewhat cluttered appearance (too zoomed out); the focus is no longer on the cyclone itself
  • Few pixels over a large geographical area, meaning lower resolution
  • Colour and hue have a slightly faded, darker and less vibrant appearance
Compromise ChocolateTrain
  • The cyclone is shown to the viewer with few unrelated cloud formations (the only non-cyclonic clouds are at the very top-right of the image)
  • The image is zoomed out to a certain degree such that:
    • An equally-sized portion of Baja California Peninsula as in the "Opposing image" is present so that it can be used as a scale reference to determine the cyclone's size
    • Unnecessary amounts of the Pacific Ocean are not included as it is neither the focus of the image nor does it convey any sense of scale (it is a monochrome, featureless expanse)
  • All spiral banding features are clearly visible
  • Many pixels over a medium-sized area, meaning high resolution (actually identical resolution to the original image)
  • Richness of colour and hue
  • The cyclone eye, though clearly visible, is slightly cloud-filled

The line through the "opposing image" is a very significant factor that contributes a poor quality to the image. It is caused as a result of the satellite SNPP's construction of images. Rather than having the black data gaps present on NASA Worldview when viewing imagery from the satellites Terra and Aqua, SNPP combines different images so as not to create gaps. Unfortunately, this creates lines straight through images where pictures taken at different times—and therefore possessing different cloud formations—are spliced together. This produces fragmented images which look, to be honest, fairly poor.

The cluttered nature of the "opposing image" also detracts from the 'viewing experience', for want of a better phrase. There are significant amounts of non-cyclonic cloud formations present in the image. Additionally, according to my calculations, the cyclone itself (incorporating the eye, central dense overcast and the spiral banding features) only takes up between 10% and 15% of the entire image area. This is in contrast to the 25% to 30% image area for the "compromise image". Having the cyclone itself occupy such little area in the "opposing image" makes it moderately unclear as to what the focus is, and it detracts from the purpose of the image being there in the first place. The need for article images to depict the subject of the image with significant size is outlined in Wikipedia's Manual of Style regarding the use of images—particularly the fourth paragraph of this section. In fact, that section also makes some more important points which are relevant in our circumstances, regarding the "opposing image". Including the point just discussed, the things it stipulates should be avoided are:

  • Depicting the specific subject matter of the image in too small a size or portion of the entire image
  • Having a cluttered image that makes the subject matter itself less distinguishable
  • Blurriness in images (this is evident on the right-hand side of the line created by SNPP, especially in the upper part of this area)

In your edit summaries, you raised the point that the "compromise image" was "zoomed in too far". Today, I had a look at some random articles in Wikipedia's featured articles list. In all, I analysed the lead images of eight articles, most of which came from the meteorology section. In each image, I calculated the approximate area occupied by the specific subject matter of the article as a percentage of the entire area. The articles I looked at were those about Earth, maple syrup, the 1980 eruption of Mount St. Helens, tornadoes, Hurricane Ioke, Typhoon Tip, Hurricane John (2006) and Tropical Storm Nicole (2010). Overall, the average percentage area was approximately 57%. As the the area that Hurricane Dora occupies in the "compromise image" is 25% to 30%, this shows that there is no problem with showing the article's subject matter in more detail by zooming in further. I also uncovered a qualitative rather than quantitative example which demonstrates my point very well, if the mathematics and statistics do not appeal to you. It is the image of Hurricane Rick in 2009. Amazingly, the hurricane is pictured in almost exactly the same spot as Hurricane Dora. It is clearly identifiable that the camera is zoomed in on Rick to an equal degree as the camera is zoomed in on Dora in the "compromise image". In fact, this image is a featured image and was Picture of the Day on 22 December 2012. This demonstrates beyond reasonable doubt that showing the cyclone zoomed in is not bad whatsoever (unless it is zoomed in to an enormous degree, which neither the original image nor the "compromise image" is).

The zoom-out that has been performed in the "compromise image" allows Baja California Peninsula to be used as a scale, and it also limits the excess of what is the unnecessary and wasted space in the cyclone's surrounds—the Pacific Ocean. In the "opposing image", most of the left side is the Pacific Ocean. This does not support the purpose of the image, nor does it provide any extra scale reference due to the lack of topographical features by which the size of the cyclone can be roughly determined. Consequently, the large amount of Pacific Ocean in the image is redundant. It is necessary to have some, of course, as the cyclone is over water, and it would be impossible and ridiculous to remove it entirely, but the "opposing image" has too much of it. The "compromise image" blends well-sized proportions of land, sea and sky to create a balanced image. The original image I uploaded, I concede, did not have enough land and was perhaps was zoomed in too far; however, if you think about it, we are meant to be talking about and therefore showing the cyclone, and not the stuff around the cyclone.

Although less of a concern, the matter of image resolution is also something which should be discussed. The "opposing image" contains 38,936,865 pixels. This is by no means bad, but it pales in comparison to the 87,100,039 pixels of the "compromise image". When taking into account the fact that the "opposing image" covers a geographical area approximately (as in, pretty rough, but I still tried to be as accurate as possible) 90% larger than the "compromise image", the "compromise image" would actually have approximately 165,490,074 pixels (ignoring significant figures) if it was the same geographic size. This translates to about 4.25 times as many pixels. This is approximately 2.06 times as many pixels in both the x and y (horizontal and vertical) axes. Now, taking into account my approximations during the productions of these numbers, this number would round down to 2. As I selected the 125-metre resolution for the "compromise image", the "opposing image" would need to be 250-metre resolution so as to create the difference in axis pixels by a factor of 2. Providing the viewer/reader with an image of higher resolution (in this case, the 125-metre one) is always better, as it allows them to use the image in potentially many more diverse ways, particularly in the field of graphic design if they so desire (provided that it is in the public domain, which, of course, both these images are). Another aspect of visual quality is colour and hue. All three images are good, but the original image and the "compromise image" have more vibrant colour and hue, whereas the "opposing image" is darker and has the effect of being a bit faded in comparison with the other two.

So, in summary, there are numerous reasons for why the so-called "compromise image" should be the one to be displayed in the 2017 Pacific hurricane season article for the purposes of showing to readers what Hurricane Dora looked like. Significant reasons include the fact that the line that cuts right through the "opposing image" from top to bottom effectively ruins it, that the "opposing image" is cluttered with non-cyclonic clouds which remove focus from the intended subject matter itself, and that there is a large amount of ocean in the "opposing image" which consequently means the cyclone occupies no more than 15% of the whole image. Other more minor, but still relevant, reasons which support the "compromise image" are that it contains quadruple the number of pixels (relative) and is therefore much higher in resolution, and that the colour and hue are richer and more vibrant, which is always advantageous. I'm hoping that this is enough to persuade you to adopt my viewpoint on this matter. I believe that the reasons I have presented here outweigh to a significant degree any points that I have thought of, and which could conceivably be thought of, which may orient one's perspective in opposition to that of my own.

Thank you in anticipation for your cooperation in this matter. ChocolateTrain (talk) 14:04, 30 June 2017 (UTC)[reply]

Retrieved from "https://en.wikipedia.org/w/index.php?title=File_talk:Dora_2017-06-26_2020Z.jpg&oldid=788272479"