I am modelling the stream delineation network from 1990 till 2020, but from USGS site, the only available data was SRTM for 2002 and 2014. How can I model it? As the DEM obtained from topographic maps are not that much accurate.
You may need to use aerial photographs and photogrammetry. SRTM, ASTER, SPOT-5 satellite DEMs are available after 2000.
I want to add that all SRTM products are derived from the mission from the year 2000, even if the products may have generated at a later time. So the data you mention from 2014 does not necessarily reflect the surface heights from 2014.
Andreas Braun is correc, SRTM is year 2000 ground. What follows are just enhanced products but the topography is the year 2000. So anywhere you can get SRTM.
There are many paths to the same end result, so are you building a hydro model or attempting to do change detection or something else? Stream delineation from DEMs are an approximation, ad dependig on the terrain, can vary considerably from actual stream courses - that doesn't matter much for hydromodels, but it may for other uses.
The world is a very large place and there are many platforms and sensors over the past 60 years, from space and the air, operated by military, civil, and commercial organizations. So the 'where' is very important, what is the smallest bounding box in latitude and longitude of your area of interest?
Why is 1990 important?
Michael John Patrick My thesis objective is to show the spatial and temporal changes of two main streams of the selected study area over the past 20years at interval of 5 years. Now, the point is if i will make the stream delineation network from topographic maps, the accuracy would be low but the DEM model from the imagery is available after year 2000.
Minza Mumtaz The 'where' ( AOI, Area of Interest ) is really important, the world is a big place. DEMs are very, very expensive to make, so the update cycles are, for global data, exactly one acquired with upgrades every five years. Our county, very wealthy, flies LIDAR every five years or so, so we have an excellent DEM, but it still would be insufficient for delineation. I recently used a 1 meter resolution from the USGS for that purpose - but doubt it will be updated for another decade. Even then using a DEM to generate a stream would be horribly off, with one exceptional geomorphology case.
And topo map contour lines of course would be even worse.
The best method is to use imagery, paying attention to time of year ( because of water levels ) - and this is why 'where' is important, if you are looking at a coastal mangrove swamp or a Himalayan mountain valley. It is easy to get a time series back to the early 2000s ( attached, PakistanStream.jpg, is a stream near Jhal Magsi, Balochistan: Pakistan lat 28.423 N, long 67.482 E ) and depending where your AOI is, I have sometimes found imagery back to the 1960's at 2 to 4 meter resolution.
These could then be georeferenced to each other, histogram adjusted, and depending on the contrast the water edges extracted, or even traced / digitized ( depends on your AOI extents, are you doing a couple kilometers, or 100s of kilometers of stream length ).
BTW, using Google MyMaps ( https://www.google.com/maps/about/mymaps/ ) you can draw your AOI as a polygon or rectangle, and save it as KML/KMZ, for folks helping you and useful for searching data catalogs, or just send as a link: https://www.google.com/maps/d/edit?mid=1cVXgJ9Idk9i_jH5RvEtKGg1cMLtUXoBe&usp=sharing ).
Michael John Patrick Thank you very much for your help. Can I contact you in future for further guidance?
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India