Hello Mr Sharma,

its relative simple. If you are using a GIS (like ArcGIS, SAGA GIS or QGIS), there should be tools like filling sinks (gaps) (Fill) and also calculating the watershed. For ArcGIS the toolbox is called: "Hydrology" (within the Spatial Analyst Tools).

In SAGA GIS: You could use the tool: Close gaps. Also the Terrain Analysis - Hydrology could be from interest.

Also check if you use the right input-format before you start.

Hope this short answer can help you.

Kind regards,

Thomas

Thanks Dear Thomas,

                                    for your reply, yes i have used fill tool of arcGIS, but after processing gaps are still there.

i am attaching here the dem, any other suggestions?

Thanks again.

Are the gaps "no data" values or "just gaps"? If these values are "no data" values then you can try to reclassyfy your dataset and setting all "no data values" to a specific value. Then try the fill-tool again. Maybe it will solve your problem.

Best regards,

Add your answer

Thomas

they are no data values, ok i will try the same.

Thanks.

Multiple ways to do:

1. Interpolate: Using interpolation methods like nearest neighbour or Krigging you can fill the gaps, but only if they are not very large holes!

2. Re-sample from other datasets, like Aster, GTOPO, or better still Cartodem (if for India)

3. Fill by a single value! 

4. Scale up, Scale Down; Re-sampling to a lower resolution and re-sampling to original resolution preserving the original spatial variance!

Thanks dear Ghosh...

Hi Sharma, 

You can use the following script in Python or Arcgis to fill NoData values in DEM. This method uses a moving window that searches for NoData values and fills them with an average.

import arcpy

import arcpy.sa

inRaster="C:/raster/DEMwithNoDatavalues.tif"

arcpy.CheckOutExtension ("Spatial")

newDEM = arcpy.sa.Con(arcpy.sa.IsNull(inRaster), arcpy.sa.FocalStatistics(inRaster, arcpy.sa.NbrRectangle(5, 5),'MEAN'), inRaster)

newDEM.save ("C:/raster/newDEM.tif")

I hope it's useful.

However, as you want to use the data in hydrological modeling, it is important to verify if the new data does not compromise the quality of the DEM for generating flows...

Mr. Sharma,

My guess is that you are refering to the CGIAR DATA:

www.cgiar-csi.org%2Fdata%2Fsrtm-90m-digital-elevation-database-v4-1

These are supposed to be Void free. The "voids" or gaps in this dataset are the ocean mask. Note that inland water are already filled with the water level elevation. By the way, nodata is set to -32768.

For hydrological modelling, water will flow into these "voids" at their river mouths.

Hope this makes sense to you!

Thanks Dr Laercio.

Fill sink operations are available in GIS, Archydro, or any of the terrain processing packages.

Good morning ,

you can start with interpolation methods like Kriging, use Aster for Re-sampling from other data sets, and Resampling at a lower resolution and re-sampling at the original resolution while preserving the original spatial variance!

good luck

You can see this video https://www.youtube.com/watch?v=zTqb6E_kkNw .... I hope you'll get the best solution from here.