Nice piece of work you have here. I have been playing a bit with it. The functionality as well as the interface are really nice. Well done.
I would like to have some swath planning capability to your application. I don't know if you have ever looked into that and if you are familiar with Earth Observation missions, although I am pretty sure you are? Is this in your near, mid term future plans for JSatTrack? I think it would provide a nice open source alternative to some rather expensive software - and, almost, be as good!
Thanks! I have been mulling over some similar ideas. Though I wasn't sure exactly what would be most useful. What are some of the basic types of swath planning capabilities do you think would be the most beneficial? Maybe constellation access plots?
from a user point of view, most useful would -basically- be to be able to: - define and attach instruments to the satellites - visualise the swath paths for each satellite instrument and for each of its modes - define a ROI on the globe and ask for intersections with satellite instrument swaths within a certain time frame This would be a good start. These items need to be refined in terms of query parameters and visualisation modes but you get the idea. The functionality is nothing new. There are already many existing tools, eg STK, SaVoir or EOLISA just to cite a few. But this would certainly the first open source version I know. When I see how much some of these software providers ask for! Having it open source would allow many EO users to add very specific functions. I have a few in mind. Let me know if I can be of any help. I have some programming experience in Java and also with WWJ. It shouldn't be too difficult to add as swaths are defined from the satellite orbit track. I started looking into your code but decided it was better to ask you first.
I think these would be great additions to the project. I have been looking into sensor types and footprint shapes since your last message. It seems a few good ones are conic, rectangular, and SAR (axe-blade). And they could each have different pointing attributes, fixed, spinning, or have a list of targets.
Thomas do you know any good references on the math/algorithms for computing the access/visibility of a satellites to a generally shaped and generally pointing sensor? It seems the problem gets quite tricky when you aren't staring straight down. The conic case might not be too bad -- though when you account for the oblateness of the Earth I imagine the equations get nice and long or you maybe a root finding type algorithm would be more efficient.
Yeah, my code is kind of messy - but it is getting cleaner as I get better with java.
I suppose it gets quite tricky when you try to do it in the general case. The good thing about those EO missions is that each sensor coverage is more or less defined based on the satellite ground track, which you already have in your application. The Satellite and its operators make it sure it is always pointing in the right direction to do so. So the bottom line is that you could take your satellite ground track and draw the footprints as in this RadarSat-2 description -figure 3 of this web page http://directory.eoportal.org/get_announce.php?an_id=7388. If you want to do it from a more sensor perspective, I suppose you would have to compute the intersection of the sensor beam 3D shape with the earth surface. May be something is available in JTS, I will check. In any case the previous method should give better results for when the swaths are defined. One way of checking this would be to do the first method quickly with one EO mission for which we have the swaths, eg ENVISAT ASAR Image mode, and compare the results with a swath planner ones, eg http://catalogues.eoportal.org/eoli.html. I will look into teh swath definitions of ENVISAT ASAR and send you some text soon.
The code doesn't look too messy -you should see mine!- and more important it works well.
Shawn, at http://earth.esa.int/object/index.cfm?fobjectid=3772&contentid=3801 you will find a table (formatting is wrong but I am sure you can work it out) giving you the swath width and incidence angles range for the different image mode beams. You can use either the swath width and one incidence angle or both incidence angles to find an approximation of the areas being imaged in each beam mode - ie IS1 to IS7.
Hey Thomas, thanks for all the links and info. I have done some reading this weekend and think I am getting a better idea of what you are thinking. Especially the last link with the java applet of the swath planner (really cool!). I also found some good math on the web about intersections of cones and ellipsoids - but I think that can be for visualization and I will follow you advice and start with the swath calculations. Though I have some questions for you.
1) The RADARSAT-2 seems to have a "blind spot" at +-10 degrees from nadir, so would you show the right and left sides as seperate swaths? This also applies if you break down the different modes (in the next question) 2) How would you envision the user defining a sensor to attach to a satellite object (in the GUI) -- I see there are many modes, would you lump them into one minimum and maximum swath look angles or would you break them down by mode and let the user choose the sensor and mode when doing the swath intersection calculations? For each mode it would seem that the data needed from the user would be the min/max incidence angle, is this correct? (or maybe a constraint on max elevation angle allowed?) 3) Can most sensors image the whole swath or just a subset of the swath at a time depending on the imaging requests. 4) Besides having a ROI for swaths are point targets also useful?
Yeah I think your last message is only visible to me as it shows up pink when I am logged in.
Sorry for the slow replies it has been a busy week for me.
Thanks for the great info though, I think I have enough to start out with and we can refine things as we go along. After I do some experimenting with a couple of ideas I will let you know how it is going and maybe I will come up with some ideas for help if I get structured enough :), or if you have any other ideas feel free to let me know.