In my expirience ASTER GDEM is global elevation data source appropriate for medium and small-scale applications. Its resolution and noise exlude it for more detailed studies. Definitely, it is one of the best global data source available for free.

In my experience, getting good elevation data from DGPS measurements is quite challenging. Thus, I'd say there are better techniques depending on the scale you are interested in and the amount of money you are willing to invest. For instance, terrestrial or airborne LiDAR may enhance your data acquisition a lot. If you have only few features you need to measure, a theodolite is probably a good option.

LiDAR would be the best choice, indeed for large scale mapping. But if you intend to do mapping at medium scale then DGPS would be ok. Another option is to go for stereo imagery/photographs. All it depends upon the scale and purpose of mapping, time as well as money available.

Personally I got LiDar images. Of course the elevation is not very accurate compared to a DGPS but you can get many other information like aspect, slope solar radiation etc

Thank you all....

I think and really progressive method is the usage by UAVs and densely overlapping image sets using the "structure from motion" procedure. You can either use fixed-wing UAVs or Copter-type-UAVs, whereas the latter will more likely produce better results. To my perspective the only limitation is the area coverage.

Most copters can cover 10 ha per flight, whereas fixed-wing UAVs have a significantly higher endurance.

I am currently validating this method for DEM genereation in the south pacific. In a month or so I can give you more information about accuracies and general results / experience.

In my experience, it depends by the extent of the study area. For small area the best solution is probably the DGPS, but you will obtain high accuracy punctual elevation data that have to be interpolated to obtain a DEM with a lower accuracy.

If the area is large, or it is not easily accessible, the best solution is probably a LiDAR that will give you a more accurate DEM. If you are interested in the submerged beach also, you can consider to use a bathymetric LiDAR that will give you a seamless DEM of emerged and submerged coast. Consider also that with the LiDAR data will came the aerial pictures, and is easy to couple to the laser sensor to a hyperspectral sensor to acquire information on vegetation and sediments.

A right person to ask this kind of question would be a land surveyor. He/she will rephrase it for you according to constrains of your area of interests (AOI) and a priori accuracy assumption. Methods to obtain a sufficiently accurate digital terrain representation of your AOI include spirit levelling, tachymetry, photogrammetry, LiDAR and InSAR. All these methods are not classified as remote sensing methods.

Almost every method has been suggested by various experts. The desired vertical accuracy is 15cm with contour interval of 0.5m, the best option avalable is Airborne LiDAR. However, the veritcal accurcay can be 70cm to 1m, any high resolution stereo images like world view/ Geoeye would suffice and eayy to generate.

Airbone LiDAR may be expensive, TLS (Terrestrial Laser Scanning) is also a relevant option !

Generally, the answer to this question depends on the accuracy you want to receive. If you asked about remote sensing data, you are searching technics for large area. In my opinion airborne laser scannig (ALS) is actually one of the best technique for it. For smaller area, mobile mapping system (MMS) mounted on on-road vehicle can be also proposed. You could also consider airborne hydrographic laser (ALH) - one of the most sophisticated technics, recently developed which provides information about bathymetry but depends also on the purity of water reservoirs.

ASTER GDEM, but it only with spatial resolution of 30 m, you may check if you need it.

free regist plus download.

In my expirience ASTER GDEM is global elevation data source appropriate for medium and small-scale applications. Its resolution and noise exlude it for more detailed studies. Definitely, it is one of the best global data source available for free.

Agree with Dr. Bakula. The ASTER GDEM has updated in 2012 (?maybe). the new version of the dataset can be utilized at 30m or larger resolution's spatial analysis.

Thank you all for your responses. In our case, the maximum elevation would be around 5 to 6 mts and its at large scale, in that case i can not go for ASTER or SRTM, i need at atleast cm or less then a meter accuracy.

Have you considered multi-overlap aerial photography? Current image matching techniques (such as semi-global matching) can produce very high quality DSMs, comparable to those from Lidar. 15cm vertical accuracy is possible, but of course you will have to fly the right camera at the right height. As previously mentioned, UAV imagery is an option, if you can get permission to fly over your area of interest.

Considering your goal in project, LIDAR or image-based technics are the best way. The cheapest way is to search data in some national repositories. Maybe there are some available. I am agree with David - photogrammetric approach is comperative to LIDAR (excluding forested areas). You can estimate then the vertical error with GSD value of photo.

I second Teja's comment above about Structure from Motion (SfM), which has successfully been used as a cheap and simple means of producing DSM, DEM, DTM, CHM.

Seek out this article for a good overview:

'Structure-from-Motion' photogrammetry: A low-cost, effective tool for geoscience applications

Hi Justin,

There are several software packages that do exactly the same as the 'SfM photogrammetry' method described in this paper ‘Structure-from-Motion’ photogrammetry: A low-cost, effective tool for geoscience applications' by M.J. Westoby et al. Try this Agisoft: http://www.agisoft.ru/.

The issue with ASTER will be its problem for relatively flatter areas like coastal areas. Its vertical resolution is not good enough to do coastal mapping. In addition 30-m is not small enough to capture some of coastal area topographic variation. Again it all depends what you would like to study. If it is not expensive, LiDAR is best.

Coastal topography variation can be captured better when the contour interval is less than one meter other wise minor variation are difficult to understand in the coarser resolution DEM

You can use Cartosat-1 stereo image (comparatively less expensive and resolution is 2.5 meter) for preparing DEM where you need few good ground control points. But LIDAR is the best one as suggested by above answers

Airborne lidar is great for detailed height measurements, but sometimes difficult to obtain. Apart from above mentioned alternatives, one could also think about using InSAR applications, e.g. from TerraSAR-X, their HDEM product (available on special request according to the produc guide) should get a vertical accuracy of 0.8 m.

NASA's ICESat has good vertical accuracy (in the order of 10 cm if I recall correctly) but its horizontal spacing may be too wide for the coastal applications and it is only available till 2010. ICESat2 is planned for 2017.