badar Posted December 12, 2011 Report Share Posted December 12, 2011 Hi, I just started a thread but this question I guess is different from that so I posted this as a new question. Whenever I try to find the cell size of some SRTM DEM tile, it is never 90m. It might be like 87.5 89 etc. Same is the case with all images like landsat cell size is never 30m MODIS is never 500m exactly. What is the reason behind this? This question has been bothering me for very long and I believe, I once knew its answer :smiley: but can't remember/find any logic/reason behind this now. This might be a silly question but any help will be highly appreciated. Quote Link to comment Share on other sites More sharing options...
pasfans01 Posted December 13, 2011 Report Share Posted December 13, 2011 based on my experiences, its caused by coordinate system transformation, like latlong to utm vice versa Quote Link to comment Share on other sites More sharing options...
Lurker Posted December 13, 2011 Report Share Posted December 13, 2011 hah, I never notice this, nice just my 2 cent though, I think its because many possibilities : 1. from the censor itself, there is a limitation for the censor to have the exact size 2. the orbital of the satellite, the way it record the earth, 3. or because the geometrical correction itself discussion is open, anyone have the better explanation? Quote Link to comment Share on other sites More sharing options...
badar Posted December 13, 2011 Author Report Share Posted December 13, 2011 @pasfans01 While performing some coordinate system transformation, I believe you have to give the cell size your self as the input. If cell size is 30 m for example, then why would some1 give non-custom cell size like 27.5 by 28 while projecting. @lurker 1. if there is some limitation of the sensor involved then no matter what non standard cell size some image has, images of that particular satellite should have the same cell size throughout all images captured. Is this the case? 2. How can we check whether orbit of the satellite is the reason of difference in resolution or no? If this really is the reason then it can be justified that cell size depends upon the distance of sensor from object being captured. orbits of satellites are elliptical so their distance from ground varies and hence the cell size. 3. This statement implies that satellite/sensor captures images with a cell size exactly equal to the mentioned one e.g. exactly equal to 30m. The cell size is changed while georeferencing, rectifying and applying geometric corrections etc. so the pixel of raw satellite imagery should depict an area on ground exactly equal to 30 m in size. Is this really the case? Quote Link to comment Share on other sites More sharing options...
Lurker Posted December 13, 2011 Report Share Posted December 13, 2011 Badar, my question to you, how do you check the exact pixel size? this is really interesting, I will dig it for more Quote Link to comment Share on other sites More sharing options...
badar Posted December 13, 2011 Author Report Share Posted December 13, 2011 Please find the attached images though I am not sure about the process of uploading images on this forum. In ArcGIS you click on the "Source" tab in the layer properties and in erdas imagine viewer you click on "i" symbol in the viewer. Here I have resampled landsat image at 30m resolution Quote Link to comment Share on other sites More sharing options...
Lurker Posted December 14, 2011 Report Share Posted December 14, 2011 it show the exact pixel as in the product description, I mean the "real" pixel size, as you have mention above It might be like 87.5 89 etc how do you get the 89 or 87,5 ? :shocked: Quote Link to comment Share on other sites More sharing options...
badar Posted December 14, 2011 Author Report Share Posted December 14, 2011 Well, here is the raw srtm file downloaded from "cigar" for reference. 1st of all the tile downloaded is in geographic coordinate system, as shown in pic below. Here the tile is not square because my dataframe coordinates are projected and layer's are geographic. cell size here is shown in degrees. Now below is the screenshot of when I try to export this tile. and I am also surprised to see that default cell size given is different than the original one (though x is same but y is different) and finally when i select my dataframe coordinates as the projection of my output file (Projected coordinates), the default option given by ArcGIS are as follows If we ignore y and only consider x, then it is about 84 in size. So does that imply that the only reason behind is the conversion from one coordinate system to another? because for MODIS also, we have to convert from sinusoidal projection (default) to UTM projection for example. and if we don't mention the cell size, the output cell size is not 500, 500 exactly. Actually I am going to present some of my work in front of professionals and I am expecting this question. And I want to be sure of my answer that what I am saying is true :undecided: Quote Link to comment Share on other sites More sharing options...
badar Posted December 14, 2011 Author Report Share Posted December 14, 2011 Ahh. I don't know why my pics are not shown in above post. You can right click on the symbol and click on the open image in new tab to view the pics. Regards Quote Link to comment Share on other sites More sharing options...
Lurker Posted December 14, 2011 Report Share Posted December 14, 2011 badar, I just edit your post, and no links to the pic, please PM the links, and I will fix your post regards, Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.