Hi Innocent,

First, I do not know at what scale you want to work... but lets say that a Landsat image can fit your needs. You can start downloading the image of your study area and perform a classification (supervised or unsupervised) of land classes, aiming at separating water bodies from the rest. Once you have the water class separated you can export this raster dataset to a vector environment, or continue working in raster environment. This is a very general indication, but as I do not know what are you trying to do is the best I can tell you.

Regards

Luis

In shallow water areas where the visibility is good (down to one secchi depth) you could use airborne laser scanning with green lasers for a mapping of the bathymetry using remote sensing.

Hi, I think that both approaches are good, satellite images or LiDAR. It depend if you have LiDAR data for your research area (Green laser pulses at 532 nm that reflects from bottom and Near-IR at 1064 reflect water surface – it is to distinguish between water surface and the bottom). The other option is chose satellite images, if you have high resolution like Ikonos, QuickBird or any kind of high resolution image could be good. The scale is an important issue in that part, because resolution has impact in your results. The next option in satellite images is Landsat, that is pretty good and the coarse resolution are MODIS, NOAA, etc. With satellite images the most ease method is an unsupervised classification (consider that your effort is the classification of water bodies). .

Hi Innocent,

besides above answers, an alternative approach can be using the ability of water to maintain a level surface in a terrain dataset. Depending on the scale of your application, you can use a SRTM 90 m, ASTER GDEM 30 m or a higher resolution DEM.

Hey,

using LANDSAT with appropriate band combinations you would have quite fast results in mapping water areas. More on band combinations you'll find here: http://web.pdx.edu/~emch/ip1/bandcombinations.html

Once you combined the image in the spectra of your interest, you can use superwised classification to extract the water class from the image. ArcGIS provide the oportunity of interactive supervised classification of images within "Image classification": http://resources.arcgis.com/en/help/main/10.1/index.html#//00nv00000002000000

Best regards,

Jewgenij

If the lakes are showing highly heterogeneous spectral characteristics, e.g., high and low sediment loads, you may also use a DEM-based approach. In most cases free data sets, such as ASTER GDEM2 or SRTM, will work. You just have to calculate the slope and subsequently extract areas where slope is very close to zero. In ArcGIS, the latter may be done using the 'reclassify' tool, for example. Afterwards you should check the results and eventually adjust the threshold value.

In many settings, you can get quite useful approximations using this technique. However, it is not sensible to whether your features actually consist of water or not. For instance, it will also delineate flat paleolake or playa-sabkha settings. This may of course be also a feature, depending on the research question.

Regards,

David

Dear Innocent

I agree with all mentioned above. There is another way to do that using water index if you want only the water bodies, but firstly you must do an atmospheric correction to get good result and when you finish you can compare with the two images before and after approach. Finally, you can use the water index to and after you got the result you can do the slicing to the image to determine the depth of the water and when you finish you can convert it to vector layer and put it in the map.

Hello everybody,

I will first and foremost appreciate you all and your wonderful contributions. I want to say, that this lake in question is believed to be a sacred one and hence no sounding likely possible, but rather the remote sensing principle. Again, all of us in the house knows that LiDAR would have be very wonderful, but in this part of world, this technology is not reachable (can someone also tell me what it will cost to carry out urban/landscape or water bodies mapping using this latest technology?).

I agree, with Luis that scale is important, but at present we are still in preparatory/planning stage and scale inclusive, thus, Luis, if you would not mind can you please suggest to us the best range to map a lake feature.

Hi Innocent,

Please send me a print preview with Google (use the satellite image please as base) and the coordinates of your lake. I will let you know what I think. One more think, please tell me what features you are needing to map and for what kind of purpose.

Regards

Luis

You can map a lake water body with remote sensing images. You may use LandSat data, which are free, to delineate the water body. Landsat band combinition

red: 0.61 - 0.69 µm

green: 0.51 - 0.60 µm

blue: 0.45 - 0.51 µm

or

VNIR: 0.76 - 0.90 µm

red: 0.61 - 0.69 µm

green: 0.51 - 0.60 µm

will provide clear indication of water body by with blue color. Please check following link for more information:

http://gdsc.nlr.nl/gdsc/en/information/earth_observation/band_combinations

Hope this helps.

Hi Luis Inostroza,

Let me have your email address where I can direct the Images and coordinates please. counting on you!

Cheers!