Merging DEM rasters?

Merging DEM rasters?

I've been struggling with this problem for awhile. Here is a curvature raster to illustrate the problem.

I'm using 10m DEM rasters from the NRCS Data Gateway for a large portion of California. The problem is some strange tiling that appears after merging, but isn't present in the original data (before merging). The grids create low points, so flow accumulation etc is drawn into them.

I've tried mosaic to new raster, create raster catalog and filter in ArcGIS 10.2, merge and build vrt in Qgis2.4 but the problem persists throughout.

Something to note is that there are apparently two sets of grids, one with regular equant shapes, and then one set where tiles overlap (on both edges of the tile). I discovered this with "build raster catalog" which creates a feature displaying the edges of the files in the catalog.

The lines are flat lows, which creates a problem trying to do any sort of slope stability or hydrologic analysis… which is the whole point of this exercise.

use BILINEAR interpolation (in environment settings) for DEM, these grids are caused by the default nearest neighbor interpolation.

This was a tough problem to track down because I had thought that the effect was at the edges of tiles when it fact they are throughout the data. You're right that the phenomena isn't in the data before mosaicking the data. The problem results from the resampling process that is inherent in mosaicking. You need to use either the cubic convolution or bilinear resampling methods when you mosaic your DEM tiles instead of the default nearest-neighbour resampling method. The phenomenon is the result of duplicated rows and columns of values during the nearest-neighbour resampling. The following is an example of a plan curvature image derived from the mosaicked tiles you provided using nearest-neighbour:

It displays the same phenomenon that you describe in the question. Nearest-neighbour resampling is most suited to applications with categorical data while continuous data, such as elevation, is better resampled using bilinear or CC resampling methods. Here is the same data after mosaicking the data using cubic convolution:

Beware, however, that bilinear resampling and cubic convolution will result in a slight smoothing of the topography (bilinear slightly less than CC).

P. Roach, Computerized Hydromechanics [Russian translation], Mir, Moscow (1980).

A. B. Veksler, D. A. Ivashintsov, and D. V. Stefanishin, Reliability and Social and Ecological Safety of Water-Development Works: Assessment and Adoption of Solutions [in Russian], Izd. JSC “VNIIG im. B. E. Vedeneeva,” St. Petersburg (2002).

A. P. Lepikhin and A. A. Titunov, “Characteristics of the construction of hydrodynamics models for sections of the Kama Reservoir most affected by human activity,” in: Modern Problems of Reservoirs and Their Spillways. Vol. 1. Hydro- and Geodynamic Processes. Chemical Composition and Quality of Water. Proc. of the Int. Sci.-Pract. Conf. (26 – 28 May 2009, Perm‘) [in Russian], Izd. PGU, Perm’ (2009), pp. 71 – 77.

S. V. Belikov and S. V. Kovalev, “Numerical investigations for solution of hydraulics problems,” Gidrotekh. Stroit., No. 8, 61 –67 (2009).

V. A. Shlychkov, “Numerical modeling of currents in the Kama Reservoir,” in: Modern Problems of Reservoirs and Their Spillways. Vol. 1. Hydro- and Geodynamic Processes. Chemical Composition and Quality of Water. Proc. of the Int. Sci.-Pract. Conf. (26 – 28 May 2009, Perm‘) [in Russian], Izd. PGU, Perm’ (2009), pp. 164 – 169.

N. P. Sidorov, I. S. Sobol’, and E. N. Gorokhov, “Simulation of the hydrodynamics of river sections and reservoirs subject to anthropogenic interference,” in: Modern Problems of Reservoirs and Their Spillways. Vol. 1. Hydro- and Geodynamic Processes. Chemical Composition and Quality of Water. Proc. of the Int. Sci.-Pract. Conf. (26 – 28 May 2009, Perm’) [in Russian], Izd. PGU, Perm’ (2009), pp. 131 – 136.

E. N. Gorokhov (ed.), Effect of Complex Excavation of Non-Metallic Construction Materials from Quarries Located in the Section of the Volga River from km 878.0 to 954.0 km and in a Section of the Oka River from its Mouth to km 58.0 on the Hydrological Regime and Navigation Conditions of the Volga and Oka Rivers: NIR Report [in Russian], Izd. NGASU (2009).

N. P. Sidorov, “Mathematical modeling of hydrodynamic failures in a cascade of earthen dams,” Privolzhskii Nauch. Zh., 8(4), 87 – 93 (2008).

N. P. Sidorov, “Analysis of flooding zones with respect to maximum water levels at hydrometeorological stations on a one-dimensional model of a river grid from experience gained by the SKIOVO on the Sura River,” Vodnoe Khoz. Rossii, No. 3, 42 – 55 (2010).

V. V. Velakov, Refinement of Methods and Procedures for Applied Numerical Modeling in the Hydraulics of Open Channels. Author’s Abstract of Doctoral Thesis [in Russian], Moscow (2005).

Merging DEM rasters? - Geographic Information Systems

Nick, you and your team did good job on making the course enjoyable, only problem I faced was having trouble downloading the tutorial 2 assignment data. Still it was a very good experience.

Good course, well structured to deliver the invaluable skills, ranging from data management to final output after processing. Good exposure to the toolbox, expecting more in the next course.

Course Overview & Data Models and Formats

This first module covers major concepts in vector and raster data models, scale, designing data tables, using vector attribute tables, and separating and joining data in order to use it more effectively in a relational database.


Nick Santos

Geospatial Applications Researcher

Текст видео

[MUSIC] Hello, everyone and welcome back. In this lesson I'm going to walk you through some concepts of raster data that we discussed last time but we'll see them in action in ARC map. We'll have a look at a couple of key types of rasters. We'll look at raster value attribute tables, raster cell alignment, and multi-band rasters. So first let's take a look at the two rasters on our screen right now. If we zoom in, we can see the digital elevation model here or DEM and I think it might take a little while for your eyes to adjust to it. But at some point it becomes pretty intuitive where the high values over here in the symbology are white and the low values are dark and we can see this sort of mountainous, dendritic river pattern coming in, where we have these high areas, and then these sloping drainage networks coming out of them. Now, digital elevation models are really important, because they underlie so many other things. And, rasters are a great simple way to represent digital elevation models. They're 3D information in a sort of, 2D format and these are often the way that we start generating the 3D formats that we use in terrain models and surface models and so we still consider a digital elevation model to be a terrain model of sorts. And if we keep zooming in, and I'll turn off the layer below it to speed this up. At some point we get to the actual cell size of the raster and we can start to see the pixels. And this is a great demonstration of why we use rasters for continuous information. And it's that they provide the sort of illusion that they are showing all of the detail of a surface. But in fact they are filled with discrete values just packed in so closely next to each other that we have an effectively continuous stream of information. But right in here all of these are just individual pixels. We can see the pixel boundaries here as the values change. This makes rasters great for anything that varies continuously across the landscape where our data actually lends Itself to continuous variation rather than sort of the vector model of these discreet polygons or something like that. So we can do hazard maps with rasters, we can do climatic models with rasters, terrain models as we're looking at right now. And so many other things lend themselves to this format. And you'll kind of know it when you see it. You'll start to get an intuitive sense for whether data should be raster or vector at its core. That said it doesn't mean that rasters can only be continuous information so this other raster right here, is discreet information, of a sort. And it's continuous in that they're trying to have a continuous representation of the landscape. But it's discreet in that the integer values in the raster don't necessarily have a relation to each other. Ten isn't more than five in this case, and 20 isn't more than ten. Instead each value in this raster encodes for a specific type of land cover. And it has a color map baked in, so in particular it's not that the raster is just being symbolized by some color map here it's that actually the color values are assigned to each value in the raster itself, so that we get something we actually kind of recognize. Where we're thinking that maybe these roads, or these red areas are roads or urban areas and that this blue looks kind of like a river to me too. So it can help us intuitively see what's in this raster which is again a land cover raster. So this makes for a great time to show about raster attribute tables which are something I haven't talked about before. Up until now raster haven't have attribute tables because they're not feature classes. So let's take a look this raster has an attribute table and I can open it. And it has an object ID field still and it has a value field and account field. And we noticed that this raster only has 15 records for a raster that covers a huge area with millions and millions of cells. So think for a second about what could be going on here. What it's doing is if I identify a value in here, I can get the color index here and let's pin the table open so you can continue to see it. So I can get the color index which is the value here. And then I can get the count. Now basically what it's doing it's giving us a record in the attribute table for every distinctive value in the raster rather than every raster cell having a attribute table which would be very prohibitive because itɽ be so many records. We have a record for every value in this discreetly valued raster. This provides a nice opportunity, though, because the values 11 and 21 don't mean anything to me for land cover. Those code for other values, and I happen to have the other values that it codes for right here in this comma separated values file. So, we can see that 11 means open water and 12 means perennial ice or snow and so on. And we can join that in just like we would with vector data in order to see what values these rasters code for. So, let's do that now. If I right click and I go to joins and relates just like we would with vector data. I go to join and I'll find that table. And select the value in the raster's attribute table here. And the value in the form table, the land cover type CSV. And all click okay, and it completes the join. And now instead of just seeing the value of the land cover, the coded value, I can actually in my attribute table have this information about what those mean. So that's where raster attribute tables are useful. It's usually with these not fully continuous rasters, these rasters with discrete values and where those values code for something that means something to us. You'll also notice something else going on here which is that I selected cells in a raster so we can do that we can select these developed areas too and create these selections. Unfortunately we can't do the same things with those selections as we can with vector data. It's more of a highlighting it for you to visually see it. We can't export the selected cells, we can't go use those only those, only those selected cells in a geoprocessing tool, or anything like that. What we'll go over later how to extract information from rasters, but it's not though the selection work flows. Okay so let's close the identify window here and collapse the table again and clear our selection. So now one thing that comes with the vector attribute tables is that shape area that tells us the area of each individual polygon. Now, since, again raster cells aren't polygons that's not necessarily a valid thing to hope for here. But what if we wanted the total area of a particular set of values here. Or of just a particular value. Since we have the attribute table here we can actually answer that question of how much open water is there. What's the area of the open water? So just like in a vector attribute table weɽ add a field and I'll call it area of land cover. And I'm going to make it a double because it could be a large number. And it pops up in the middle here at the end of the original attribute table, not with the joined values. And I'm going to go to field calculator. And think for a second how you would find the area of a raster. Basically, we need to know how many cells we have and multiply it by the area of the cells, right? So, in this case, we can find the area of the cells, so let's cancel out for a second, and lets go to the land cover layer here, go to properties, and we can see that the cell size is 30 by 30, so it's 30 meters to each side. So if we go back to the field calculator, we can put in the count here. And then put in multiply it by the area, which is 30 by 30. So really what we're doing is we're multiplying 30 by 30 to get the area of one cell and then multiply it by the count to get the area of all of those cells. And what it's going to do is run for that selected row, and it gives us that area of that set of cells here. So we have five billion square meters of open water. Okay now let's take a look at that cell alignment problem I mentioned last time. And let's zoom to a particular spot here. And we can see once we zoom in to the rasters, they're different cell sizes and they're different cell alignments. So the land cover raster is a 30 by 30 raster but the digital elevation model is a 10 by 10 approximately raster. So with these different cell sizes we get different cell alignments and already we can see that their slightly off, if this looks like it's one pixel here and then we have these pixels overlaying it. Imagine if we needed to combine these rasters, weɽ have a problem. So let's just look at this a little closer. I'm going to bring up the image analysis window, I'll pin that here. And I'll select the top and I'm going to use the swipe tool and I'll go over here and that lets me kind of turn off the top layer and show what's below it. So if we take a close look we can see that once we get to that bigger cell in the land cover raster, we're still not quite done with these other cells. So, these cells right here, are touching that cell. So, we have three cells and then another three cells. So, we have six cells touching it and, then seven and eight cells touching it, not including the null and then, a ninth cell touching it. And we have this not quite aligned edge here. So it doesn't match up completely over here, so we have ambiguity in how to choose which cell value to assign to which other cell value. If we were trying to, say, add these together or something, if I was trying to merge the values in the digital elevation model with the land cover and some sort of model and use 30 meters weɽ need to decide some set of rules for how the digital elevation model's values get applied at that larger cell size. Most commonly it's either an average or it's whichever one is most dominant or it's whichever one is at the center of the target cell so whichever one would be at the center of the land cover cell. Okay, the last thing I want to do is I want to show you a multi-band raster so if we switch data frames here, imagery is a multi band raster. And we can see that over on the left here. We have band 1, band 2, and band 3 in this one raster. And if we go to properties and symbology. We can assign those bands to different channels. So we can actually view red light as blue, and blue light as green, if we want to. Which, has valid use cases. But, for now, just see that we can select any of the bands in this multi-band raster. There's actually a fourth band, which is near infra-red light, in this case, that we assign to make it visible. So, we could say, take this band in this raster, and display it as red, green, or blue light on our screen. And when we display light captured by a sensor as red light, as red, and light capture by sensor as green light as green, and light captured by a blue sensor as blue, we get, visible imagery as we expect, but we can start to play with this to take multiple strings of information and make them visible in a multi-band raster. But really mostly what I want you to take away from this is that there's one raster, that has multiple layers of information built inside of it. And that we can then take those and display them in different ways, but that they're available as different streams of information for analysis as well. This might be confusing still but we'll talk about imagery in a later class, it should become much clearer. If we were to zoom in here, weɽ still see the cell size here but we don't see three streams of information, we see one because our eyes see red, green, and blue light combined so when it puts them out appropriately as red, green, and blue light we just see things we normally see. Okay, that's it for this lecture. In this lecture we went through some characteristics of rasters, from digital elevation models as continuous rasters to land covered data set as a discrete raster, and we looked at raster attribute tables and raster cell sizes and overlaps, and then at multiband rasters as well. I hope that helps you better conceptualize what raster data is and some potential uses for it. See you next time.

Merging DEM rasters? - Geographic Information Systems

Management Information Systems 2000: GIS and Remote Sensing

Author(s): U. Westman, K. Ottosson & Y. Ronning

Pages: 10 Price: Free (open access)

Size: 945 kb Copyright: WIT Press

DOI: 10.2495/MIS000001

Author(s): J. Counsell & N. Bates-Brkljac

Pages: 10 Price: Free (open access)

Size: 1,420 kb Copyright: WIT Press

DOI: 10.2495/MIS000011

Author(s): M. Bayoumi & S.T. Jerman

Pages: 13 Price: Free (open access)

Size: 1,269 kb Copyright: WIT Press

DOI: 10.2495/MIS000021

Author(s): A.L. Naja, M. Chame & S.A. de Miranda Chaves

Pages: 9 Price: Free (open access)

Size: 865 kb Copyright: WIT Press

DOI: 10.2495/MIS000031

Author(s): L. Novak, J. Stejfa, V. Dobef & M. Kuras

Pages: 8 Price: Free (open access)

Size: 730 kb Copyright: WIT Press

DOI: 10.2495/MIS000041

Author(s): J. Buzolic, N. Mladineo & S. Knezic

Pages: 10 Price: Free (open access)

Size: 1,033 kb Copyright: WIT Press

DOI: 10.2495/MIS000051

Author(s): S.D. Farrier, T.M. Slow, A.R. Barker & K. Maddocks

Pages: 10 Price: Free (open access)

Size: 1,426 kb Copyright: WIT Press

DOI: 10.2495/MIS000061

Author(s): E. Fernandez, A. Rescia, P. Aguilera, H. Castro, M.F. Schmitz & F.D. Pineda

Pages: 15 Price: Free (open access)

Size: 1,504 kb Copyright: WIT Press

DOI: 10.2495/MIS000071

Author(s): F. Rossi, M. Folino & F. Lamberti

Pages: 11 Price: Free (open access)

Size: 1,443 kb Copyright: WIT Press

DOI: 10.2495/MIS000081

Author(s): J.J. de Melo & P. Santana

Pages: 9 Price: Free (open access)

Size: 1,028 kb Copyright: WIT Press

DOI: 10.2495/MIS000091

Author(s): A.F. Shakweer & A.A. Darwish

Pages: 10 Price: Free (open access)

Size: 1,223 kb Copyright: WIT Press

DOI: 10.2495/MIS000101

Author(s): C. Alvarez, J. Mateu, T. Sanfeliu & M.M. Jordan

Pages: 11 Price: Free (open access)

Size: 882 kb Copyright: WIT Press

DOI: 10.2495/MIS000111

Author(s): J. Lourengo

Pages: 10 Price: Free (open access)

Size: 994 kb Copyright: WIT Press

DOI: 10.2495/MIS000121

Author(s): M. Campos, R. Silva, J. de la Luz & R. Cerame

Pages: 10 Price: Free (open access)

Size: 2,439 kb Copyright: WIT Press

DOI: 10.2495/MIS000131

Author(s): B. Krawiec & D. Markiewska-Krawiec

Pages: 9 Price: Free (open access)

Size: 691 kb Copyright: WIT Press

DOI: 10.2495/MIS000141

Author(s): D.G. Goulias & K.G. Goulias

Pages: 11 Price: Free (open access)

Size: 1,197 kb Copyright: WIT Press

DOI: 10.2495/MIS000151

Author(s): N.C. Babic

Pages: 8 Price: Free (open access)

Size: 970 kb Copyright: WIT Press

DOI: 10.2495/MIS000161

Author(s): F.J. Tapiador & J.L. Casanova

Pages: 9 Price: Free (open access)

Size: 827 kb Copyright: WIT Press

DOI: 10.2495/MIS000171

Author(s): J.E. Dunn & L. Duncan

Pages: 10 Price: Free (open access)

Size: 886 kb Copyright: WIT Press

DOI: 10.2495/MIS000181

Author(s): J. Garus & B. Zak

Pages: 8 Price: Free (open access)

Size: 789 kb Copyright: WIT Press

DOI: 10.2495/MIS000191

Author(s): I.T. Kristensen & P.E. Larsen

Pages: 10 Price: Free (open access)

Size: 1,055 kb Copyright: WIT Press

DOI: 10.2495/MIS000201

Author(s): D. Cluis, E. Quentin & G. Benie

Pages: 10 Price: Free (open access)

Size: 939 kb Copyright: WIT Press

DOI: 10.2495/MIS000211

Author(s): P. Juan, J. Mateu, C. Antolin & C. Ano

Pages: 10 Price: Free (open access)

Size: 848 kb Copyright: WIT Press

DOI: 10.2495/MIS000221

Author(s): S. Jassim, H. Du & M.F. Obatusin

Pages: 10 Price: Free (open access)

Size: 1,217 kb Copyright: WIT Press

DOI: 10.2495/MIS000231

Author(s): P.L. Aguilar, P. Martinez Cobo & R.M. Perez

Pages: 10 Price: Free (open access)

Size: 818 kb Copyright: WIT Press

DOI: 10.2495/MIS000241

Author(s): G.K. Adam

Pages: 7 Price: Free (open access)

Size: 794 kb Copyright: WIT Press

DOI: 10.2495/MIS000251

Author(s): D. Anderson

Pages: 7 Price: Free (open access)

Size: 608 kb Copyright: WIT Press

DOI: 10.2495/MIS000261

Author(s): J. Biggam & A. Hogarth

Pages: 8 Price: Free (open access)

Size: 812 kb Copyright: WIT Press

DOI: 10.2495/MIS000271

Author(s): A. Cammelli & E. Fameli

Pages: 11 Price: Free (open access)

Size: 1,370 kb Copyright: WIT Press

DOI: 10.2495/MIS000281

Author(s): A. Busatti & R. Paoluzzi

Pages: 9 Price: Free (open access)

Size: 1,124 kb Copyright: WIT Press

DOI: 10.2495/MIS000291

Author(s): D.T. Kincaid, R. Stalter, E.E. Lamont

Pages: 9 Price: Free (open access)

Size: 769 kb Copyright: WIT Press

DOI: 10.2495/MIS000301

Author(s): N. Posada & D. Sol

Pages: 10 Price: Free (open access)

Size: 1,183 kb Copyright: WIT Press

DOI: 10.2495/MIS000311

Author(s): J. Jaafar, G. Priestnall & P.M. Mather

Pages: 9 Price: Free (open access)

Size: 810 kb Copyright: WIT Press

DOI: 10.2495/MIS000321

Author(s): N. Dedios Mimbela, M.S. De la Orden, A.G. Ferrer-Forms & J.I. Quispe Becerra

Pages: 10 Price: Free (open access)

Size: 1,090 kb Copyright: WIT Press

DOI: 10.2495/MIS000331

Author(s): H.-U. Wetzel, S. Roessner, & A. Sarnagoev

Pages: 12 Price: Free (open access)

Size: 1,723 kb Copyright: WIT Press

DOI: 10.2495/MIS000341

Author(s): A. Sole, A. Crisci & G. Spadino Pippa

Pages: 13 Price: Free (open access)

Size: 1,428 kb Copyright: WIT Press

DOI: 10.2495/MIS000351

Author(s): M.T. Shalaby & A.A. Darwish

Pages: 11 Price: Free (open access)

Size: 1,182 kb Copyright: WIT Press

DOI: 10.2495/MIS000361

Author(s): J. de Vente & J.C.J.H. Aerts

Pages: 11 Price: Free (open access)

Size: 1,079 kb Copyright: WIT Press

DOI: 10.2495/MIS000371

Author(s): G. Deplane, L. Boggio, M. Campagna, A. De Montis & I. Onnis

Pages: 9 Price: Free (open access)

Size: 1,245 kb Copyright: WIT Press

DOI: 10.2495/MIS000381

Author(s): J. Yoon

Pages: 9 Price: Free (open access)

Size: 1,235 kb Copyright: WIT Press

DOI: 10.2495/MIS000391

Author(s): B. Jaimes, F.V. Vidal & V.M.V. Vidal

Pages: 10 Price: Free (open access)

Size: 840 kb Copyright: WIT Press

DOI: 10.2495/MIS000401

Author(s): J. Balicki & Z. Kitowski

Pages: 10 Price: Free (open access)

Size: 783 kb Copyright: WIT Press

DOI: 10.2495/MIS000411

Author(s): J. Ceuppens, D. Raes & M. Sarr

Pages: 9 Price: Free (open access)

Size: 1,062 kb Copyright: WIT Press

DOI: 10.2495/MIS000421

Author(s): J. Garus & B. Zak

Pages: 7 Price: Free (open access)

Size: 650 kb Copyright: WIT Press

DOI: 10.2495/MIS000431

Author(s): K. Nakata, M. Takei, T. Nakane G. Maxwell & D. Torpie

View Raster Properties

Every ENVIRaster object has a set of common properties that you can view but not edit. In the ENVIRaster help topic, these properties are marked as "Get."

Many of the same properties are also marked as "Init" (short for initialize), meaning that you can set values for these properties when you first initialize the object, but not afterward. You can only set these properties when creating a new ENVIRaster object, not when opening an existing file as an ENVIRaster object as in the example below.

Copy and paste the following code into the IDL command line:

Printing the raster's properties gives you some details about the raster, namely:

  • It has four bands.
  • Its pixel dimensions are 1024 x 1024.
  • Its interleave type is band-sequential (BSQ).
  • Its data type is unsigned integer ("uint").

To see the value of a specific property such as the number of bands, type the following (case does not matter):

Try changing the number of bands:

Again, ENVIRaster properties can only be viewed but not changed. Some properties for other ENVI API objects can be set after they are first initialized. These are marked as "Set" in the associated help topics.

You can also use ENVIRasterPropertiesTask to retrieve raster properties within an image-processing script, then create variables from the various properties.

Keywords allow you to set additional properties on raster objects when you first intialize them. ENVIRaster has keywords such as DATA_IGNORE_VALUE, INHERITS_FROM, and ERROR. Here is an example of setting a data value to ignore on a new raster object:

While properties let you view the attributes of a raster object, methods allow you to perform actions on the raster. Methods are referenced by their name with double colons appended to the front, for example, ENVIRaster::GetData. Refer to the list of methods that are available for ENVIRaster.

How to combine elevation values from two separate rasters?

I have two separate elevation rasters. One is a standard DEM and the other is a tilted planar surface with elevation values ranging between -97 and 97m. I would like to combine the elevation values together, adding elevation from the tilted surface to the real-world elevation value. Where a real-world elevation in a specific cell is 310m, the same cell on the tilted surface raster would have an elevation of -45, thus making my new elevation value for this cell 265. How do I do this either on ArcGIS or QGIS?

It's pretty straightforward raster math which you should be able to handle with raster functions in ArcGIS. Create a function that adds the two rasters together. You will get more accurate results if you assure that they are both in the same projection/coordinate system and have the same cell size.

My problem is that I'm not very familiar with raster math. I wouldn't know where to start.

Both layers are in the same projection and have the same cell size.

I know you asked for Arc or Q, but the easiest way I know to do this is in Manifold's Release 8 (the precedessor to the current edition), which has a nifty surface transform dialog. Here's the illustrated topic. If your standard DEM is called "DEM" and your tilted surface is called "Tilted" you just enter [DEM] + [Tilted]. Hard to get wrong.

The Release 8 user manual illustrations are all pre-Windows 10, so they look dated, but the internals are more modern than either Arc or Q. For example, the surface transform dialog has many functions GPU-parallelized, so it works great.

You could do this in Release 9 as well, but the process is more intricate and requires more explaining to allow for different raster sizes, etc. If you are doing this for very large surfaces, 9 is the way to go.

If you have download links for your data Iɽ be happy to take a look.

Why store data as a raster?

Sometimes you don't have the choice of storing your data as a raster for example, imagery is only available as a raster. However, there are many other features (such as points) and measurements (such as rainfall) that could be stored as either a raster or a feature (vector) data type.

The advantages of storing your data as a raster are as follows:

  • A simple data structure—A matrix of cells with values representing a coordinate and sometimes linked to an attribute table
  • A powerful format for advanced spatial and statistical analysis
  • The ability to represent continuous surfaces and perform surface analysis
  • The ability to uniformly store points, lines, polygons, and surfaces
  • The ability to perform fast overlays with complex datasets

There are other considerations for storing your data as a raster that may convince you to use a vector-based storage option. For example:

  • There can be spatial inaccuracies due to the limits imposed by the raster dataset cell dimensions.
  • Raster datasets are potentially very large. Resolution increases as the size of the cell decreases however, normally cost also increases in both disk space and processing speeds. For a given area, changing cells to one-half the current size requires as much as four times the storage space, depending on the type of data and storage techniques used.

Merging DEM rasters? - Geographic Information Systems

5.A.i.b. Derivation of the 30" DEM from Source Materials

In 1994, NGDC (now NCEI) and DMA jointly designed a 30" DEM which DMA would contribute to NCEI for GLOBE. In the original design, a 10x10 array of 3" DTED Level 1 grid cells was processed to determine the minimum, maximum, and mean of 3" values in each available 30" GLOBE grid cell. The data were restructured at NCEI for more convenient processing in raster geographic information systems (GIS), and released to the public as GLOBE Prototype Version 0.1 in 1995.

NIMA added a discrete (spot) 3" value to the data collection. This compilation is available at NIMA's Web site at the time of writing this document, and also on CD-ROM as GLOBE Prototype Version 0.5. The latter version has the files restructured for greater ease of use in a GIS.

Graphic describing georeferencing and sampling for DTED Level 0 discrete (spot) data (source/lineage category 1).

In addition, DMA had contributed coverage for most of the United States from an early precursor of DTED to USGS for public distribution. USGS calls these data "USGS 3 arc-second data." USGS resampled these data to 30" by nearest-neighbor techniques, for incorporation into GTOPO30. These data form GLOBE 1.0 source/lineage category 5.

Graphic describing georeferencing and sampling for DMA/USGS 3 arc-second data to 30" for the U.S.A. (source/lineage category 5).

Similarly, DMA provided 30" grids for the conterminous U.S. (from an early version of DTED) for public distribution by NGDC (now NCEI) in the early 1980s. This 30" DEM included mean and spot (nearest- neighbor from 3") values. The spot data form GLOBE 1.0 source/lineage category 4.

Graphic describing georeferencing and sampling for DMA/USGS 30" spot data for the conterminous U.S.A. and vicinity (source/lineage category 4).

As noted previously, USGS developed a 30" global DEM, called GTOPO30. GTOPO30 development involved specific groups assembling data for the different continents. Decisions made by these groups resulted in the following resampling methods:

  • For Africa (the first continent attempted), the resampling of 3" data used a &ldquobreakline&rdquo approach that favored ridges and valleys (Gesch and Larson, 1996). This was done to best fit with the ANUDEM- based methods (Hutchinson, 1989, 1996 Danielson, 1996) used for Digital Chart of the World gridding for Africa.

Graphic describing georeferencing and sampling for DTED for Africa (source/lineage category 6).

  • For Eurasia, the resampling consisted of computing median values of non-oceanic locations within each 30" arc-second grid cell.

Graphic describing georeferencing and sampling for DTED for Eurasia (source/lineage category 2).

  • For the Americas, the resampling consisted of taking a ("nearest-neighbor") 3" value nearest the center of each respective 30" grid cell. Due to the georeferencing of the 30" GLOBE grid compared to that of 3" DTED Level 1 data, there is a 3" DTED Level 1 grid cell-centered directly at the center of a 30" GLOBE cell. That value was used in the Americas.

Graphic describing georeferencing and sampling for DTED for the Americas (source/lineage category 3).

  • The DEMs for Eurasia and Africa were mosaicked along 39 o N latitude, and 59 o E longitude. The data were linearly blended along a 2-degree-wide zone centered along these lines. Thus at 40 o N, median derivations were used, at 38 o N (west of 58 o E) breakline methods were used exclusively, and at 39 o N (west of 59 o E) 50% weighting of both of these methods was used. This blending is category 7 in the source/lineage map.

Thus data originally from DTED sources have been contributed to GLOBE directly from NIMA. In addition, data were previously contributed by DMA for public distribution to USGS and NGDC (now NCEI) at various times during the past 20 years. The source map shown in Section 11.E shows where different versions of these data were used in GLOBE Version 1.0.

Some best practices for working with DEMs

Most of us use digital elevation models (DEMs) which are raster data sets that represent a continuous elevation surface in which each cell represents the elevation at its location. DEM data are typically available in tiles that are sized to balance ease of data sharing with coverage so that those who need several tiles to cover their area of interest are not overburdened with extensive post-download data processing.

We are frequently asked a variety of questions about managing DEM datasets and how to derive other datasets like hillshade or slope. Here are some of the best practices that have come from these discussions and our thoughts on them:

Manage DEMs in their native geographic coordinate system as opposed to a projected coordinate system

Always strive to obtain and maintain a “gold” copy or master of the DEM dataset — this is the data in the geographic coordinate system, for example, WGS84 or NAD83, that was used when the data were originally collected. In terms of data fidelity, this is the most accurate representation of the elevation model you have because it will not have undergone transformations, like projection, which can degrade the data for further display or analysis.

Naming DEMs and derived datasets

I include either “_ft” or “_m” at the end of my DEM’s name to indicate the units of elevation. I do the same for contours that I derive from my DEMs. This is important to know because if the elevation units are different from the linear (x,y) units, you will likely have to use the z factor to make sure your derived rasters (hillshade, slope, etc.) are correct — see the blog entry “Setting the Z Factor Parameter Correctly” for more on this.

I use “_per” or “_deg” for my slope output to indicate whether the slope is expressed in degrees or as a percentage.

Avoid projecting DEMs that already use a projected coordinate system

When a raster dataset is projected,it is distorted, sometimes greatly, and if the original dataset is deleted, some data are lost. For example, for a raster DEM that has been projected into UTM Zone 11 North, each of the northern-most cells contain elevation values that are a result of averaging/sampling a run of several cells from a row in the original dataset. If you were to project that dataset into a geographic coordinate system, the run of cells representing one northern edge UTM cell would be stretched into an evenly averaged range of elevation values.

In other words, it is only possible to maintain fidelity in the DEM data when the data are projected from a geographic to a projected coordinate system. An equivalent of the original data can never be obtained by additional projecting the data back – some of the accuracy in the measurement will always be lost.

Mosaic DEM tiles first before projecting or producing derivative datasets

If your area of interest covers more than one DEM tile, use the Mosaic tool to combine the tiles first before producing other datasets and projecting them, otherwise you will introduce error into the derived datasets. If you want to test this, try first projecting a DEM dataset and then producing a hillshade – you’ll typically see stripes in the resulting hillshade. On the other hand, if you first produce the hillshade from the original unprojected DEM and then project the hillshade, the result will look right. The same goes for all other derived products including slope and aspect datasets and more complex products, like watersheds or flow analyses that are the result of some fo the hydrographic analysis tools.

Publishing and sharing DEM datasets

Last, if you are a producer/publisher of DEM rasters for your area, please do not publish onlya projected edition of your data. Doing so severely limits what your customers can do with your data. First and foremost, publish your DEM data in the original geographic coordinate system so that is appropriate for general use. Then, if there is a standard projected coordinate system in use for your area, and you want to save your customers some time, additionally publish derived products like combined DEM tiles, hillshades, slope datasets, and so on, that you have created by first generating the derived raster data THEN projecting it.

Merging DEM rasters? - Geographic Information Systems

Thematic and continuous rasters may be displayed as layers along with other geographic data on your map.

While the structure of raster data is simple, it is exceptionally useful for a wide range of applications. In ArcGIS Explorer the uses of raster data as layers may be categorized as follows:

  • Rasters as basemaps
  • A common use of raster data in a GIS is as a background display for other feature layers. For example, orthophotographs displayed underneath other layers provide the map user with confidence that map layers are spatially aligned and represent real objects as well as additional information. Three main sources of raster basemaps are orthophotos from aerial photography, satellite imagery, and scanned maps.
    Below is a raster used as a basemap for road data.
  • Rasters as surface maps
  • Rasters are well suited for representing data that changes continuously across a landscape (surface). They provide an effective method of storing the continuity as a surface. They also provide a regularly spaced representation of surfaces. Elevation values measured from the earth's surface are the most common application of surface maps, but other values, such as rainfall, temperature, concentration, and population density, can also define surfaces that can be spatially analyzed.
    The raster below displays elevation—using green to show lower elevation and red, pink, and white cells to show higher elevation.
  • Rasters as thematic maps
  • Rasters representing thematic data can be derived from analyzing other data. A common analysis application is classifying a satellite image by land-cover categories. Basically, this activity groups the values of multispectral data into classes (such as vegetation type) and assigns a categorical value. Thematic maps can also result from geoprocessing operations that combine data from various sources such as vector, raster, and terrain data. For example, you can process data through a geoprocessing model to create a raster dataset that maps suitability for a specific activity.
    Below is an example of a classified raster dataset showing land use. Agriculture is respresented in brown, water in blue, bare ground in yellow, a variety of deciduous and non-deciduous trees in shades of green, and urban/developed land in gray.
To add raster data to the map

On the Home tab, in the Map group, click Add Content and then click Raster Data. and browse for an image file to add.

ArcGIS Explorer supports the display of many raster formats, including: Imagine image (.img), bitmap (.bmp), JPEG (.jpg, .jpeg), Portable Network Graphics (.png), Graphics Interchange Format (.gif), Tagged Image File Format (.tif, .tiff), ARC/INFO and Space Imaging BIL (.bil), ARC/INFO and Space Imaging BIP (.bip), ARC/INFO and Space Imaging BSQ (.bsq), DTED Level 0-2 (.dted), ERDAS 7.5 LAN (.lan), ERDAS 7.5 GIS (.gis), JP2 (.jp2), MrSID (.sid), RAW (.raw), NTIF (.ntf), USGS ASCII DEM (.dem), X11 Pixmap (.xpm), PC Raster (.map), PCI Geomatics Database File (.pix), JPC (.jpc), J2C (.j2c), J2K (.j2k), HDF (.hdf), BSB (.kap), Raster Product Format RPF, CIB, CADRG (.toc), DIGEST ASRP & USRP (.img).

Georeference a raster

If the raster has a coordinate system defined it will be drawn on the map. If the coordinate system is undefined you will be prompted to georeference the raster. If you choose to georeference a raster you will see the following dialog:

Pan and zoom the map to the geographic location where you want the raster to display. Click the Fit to Display button to move the raster closer to the desired location. Use the Get position button and click a location on the raster. After clicking, the raster will turn off, allowing you to see and click the corresponding position on the map. Repeat this process adding a minimum of 3 control points. Once you have finished adding control points, click Georeference to align the raster with the map at the desired location.

Tip: You can also adjust rasters that already have a coordinate system defined by clicking the Georeference button on the Raster Layers - Tools tab.