I assume you need to measure length in order to calculate area. If so, certainly there must be techniques that can be applied to satellite images that will calculate the area of any landmass. Or perhaps I don't understand what you men here?

No, I have area, and perimeter, I need island length, the longest line on the island. However, these islands have shapes that I'm not sure how to measure the longest axis (if that would be the correct term).

for example, the island with the upside-down V shape... do I take into account both "arms" of the V shape, or just the "larger, longer" one. The same thing is about width.

Perhaps the most important factor in your study is methodological consistency. This of course assumes that there are no already established, accepted methods followed in your field of study. In the case of an island shaped like a V, you could measure the longest arm and use that as the "length," or you could take the average of the two arm measurements as the length. Decide what method of measurement makes the most sense for your particular research problem, define that method of measurement, and use it consistently.

I'm not a geographer. You should probably talk to a geographer to get a definitive answer on this.

Hi Zorana,

I like Michael's answer, and Thierry's proposal for a way to operationalize it in a specific measure. The choice of measure should depend on what you want to use the shape analysis for. "width" and "length" are just words, and very bad descriptors of weird shapes like your islands. Perhaps you need some more descriptive term, like "the longest linear distance measured across the island without the line crossing any sea" (and than "width" could be the lonest linear distance perpendicular to the former one), for example. Whether you use this type of measure or Thierry's ellips, really depends on what you are going to use it for.For example, imagine you are aiming to say something about connectedness between local populations on islands with different shapes. This would determine the way in which you are going to define the shape; simple "lenth" and "width" would have very limited value here, but you would need a more descriptive measure to, for example, describe extent of "fingery-ness" (how descriptive! ;-) of the island. By the way, I found one link to a paper that seems useful in this context: http://graphics.stanford.edu/courses/cs164-09-spring/Handouts/paper_shape_spaces_imm403.pdf. The reason why I felt called to respond to this question is a limited experience I have with the analysis of insect color patterns, where we also often have shapes that don't have an obvious width or length, and need to be described using landmarks and consistent measurements using the landmarks.

Hi!

I had to measure the diameter of dolines for my thesis and I think for measuring any areal landform one could use the same methods. So I recommend this paper: Bondesan, A., Meneghel, M., Sauro, U., 1992: Morphometric Analysis of Dolines, International Journal of Speleology 21, 1-55 (http://scholarcommons.usf.edu/ijs/vol21/iss1/1/).

It provides the guiding principles for measuring length, diameter, perimeter, area etc. of various shapes including V shape (of Molat island).

Hello, Zorana,

I think that Michael was on to something in suggesting the potential value of consulting with a geographer on this. Or, you could at least consider how to answer the question from a geographical perspective (which would also address Michael's point about methodological consistency). It seems to me that considering island dimensions in terms of longitudinal and latitudinal orientation or, alternatively (and in a slightly more fine-grained manner), in reference to the eight principal compass points (i.e., N, NE, E, SE, S, SW, W, and NW), would be both relatively straightforward to determine, while at the same time being methodologically consistent. So, for example, you could describe a hypothetical island in terms of its greatest linear distance in terms of longitudinal dimensions as being along a NW to SE orientation, while greatest linear distance in latitudinal dimensions could be characterized as being along a W to E orientation.

hi thierry,

thanks for this. to be honest, I am not an expert on fourrier analyses, but I am aware it is used to describe complex shapes. I know it has, for example, been used to establish the similarity of complex patterns on the wings of the peppered moth (Biston betularia) and its backgrounds. so indeed it seems an excellent way to quantify shapes of islands as well. I hope Zorana is helped with this.

Thank you all for your answers. They are really helpfull. I will try to research some of the methods here, and definetly try to find a geographer.

I'm tryin to compare island shapes with floristic biodiversity for som islands in the adriatic. The thing is that for some factors I need the parameters mentioned here. I already have calculated shape index and fractal dimension.

Thank you again!

I'lll get back whith the results :)

Hello Zorana

I am doing something similar but am doing this in a GIS as that is my background. I have a database of 1800 islands in the Pacific and we have classified all islands into different risk factors based on climate change and lithology. We have also used shape as a factor and for shape I calculated circularity (how close is the shape of an island to a perfect circle). Probably the best way to find the longest length of an island is to fit a circle, so find the smallest circle within which an island can fit. The diameter of this circle then gives the length. I think this can work.

Email me on [email protected] if you want to discuss this further.

Lalit

Dear Lalit,

Thank you very much. I think this is a good way of seeing it.

I will look into this solution and probably contact you.

Thank you again

Regards

Zorana