POST
|
Hi Nicholas, In regards to the most recent steps you have outlined: [1] The order that you add and remove the primary rasters shouldn't matter. Deleting, defining, and rebuilding the overviews is going to have a much greater impact on anyone that is potentially using the mosaic dataset. Also, if you happen to have a service running off of the mosaic dataset, I would probably stop it. I rarely work on the service side but some changes to a mosaic dataset are not possible if a service is using it. [2] Did you rebuild the footprints on the original primary rasters? If not, then there is no reason to do this with the new rasters. Footprints are automatically built when adding rasters and automatically deleted when removing rasters. Unless you need to modify the default footprints generated by the software, then you shouldn't need to rebuild the footprints. [3] I prefer the manual selection as Larry demonstrated but I would try the settings in the screenshot below. See my comment on the "Update Cell Size Ranges" option in my previous post. [4] and [5] I would deviate from these steps in my initial testing. I would first try just using the Build Overviews tool with the settings in the screenshot below. I don't think you will need to use a query since the affected overviews have been marked while removing the primary rasters, but only testing will confirm this for sure. Since you have previously defined satisfactory overview levels, then regenerating only the affected overviews may be a more efficient approach than deleting, defining, and rebuilding all overviews. Of course, if the results are not satisfactory, you can always delete, define, and rebuild all the overviews.
... View more
06-03-2015
06:24 AM
|
1
|
3
|
427
|
POST
|
Thanks for the additional details. Since you did some customizing to various aspects of the mosaic dataset, I would stay away from the Synchronize tool. Something similar to what Larry suggested should meet your needs. Here is a general approach that I think will work but you should definitely test on something other than your production mosaic dataset. 1. Use the Remove Rasters From Mosaic Dataset tool to remove the rasters that need to be replaced. You can use a query in the tool or a selection in the Footprint attribute table to do this. 2. Use the Add Rasters to Mosaic Dataset tool to add the new rasters. I would uncheck the the "Update Cell Size Ranges" option. If you leave this option enabled in the tool, you risk the software recalculating all of your MinPS and MaxPS values for the entire mosaic dataset to something it thinks is best. After the raster are added, use a selection and the Field Calculator in the Footprint attribute table to add the desired MinPS and MaxPS values for the new rasters. 3. Verify the new rasters are displaying to your satisfaction. 4. If the new rasters represent the same extent of the old rasters, then you should be able to run the Build Overviews tool with the "Generate Overveiws" option and the "Regenerate Stale Overview Images Only" option enabled. Be sure to disable "Define Missing Overview Tiles." If the new rasters cover different extents than the old rasters, then you will probably want to remove the existing overviews and Define and Build overviews again.
... View more
06-02-2015
05:40 AM
|
1
|
9
|
1585
|
POST
|
Hi Nicholas, This is a great question and unfortunately, the available documentation does not do a good job of prescribing an approach for specific situations. This is likely due to the incredible variability in mosaic dataset implementations and the needs/objectives of users. As usual, there are probably a few different ways you could go about the update process and achieve favorable results and in my experience, setting up a test mosaic dataset and working out a proof of concept based on the documentation and trial-and-error testing has been the best approach. What you don't want to do is test anything out on a mosaic dataset that contains 1,000 tiffs and has been configured to your specifications. In other words, you want to have confirmed that the approach you have chosen will actually do what you need it to do for your specific mosaic dataset implementation. Many things you do to mosaic datasets, such as using the Synchronize tool, are not reversible so investing time and effort into testing usually saves you time and effort in the long run. Based on the info you have provided, your implementation sounds pretty straightforward but I do have a few specific questions: 1. Did you configure the cell size ranges (i.e., MinPS and MaxPS) or allow the software to do it? 2. Did you configure anything in the Define Overviews tool or accept the defaults? 3. Do you intend to update the affected overviews in addition to replacing the source images? 4. Are you OK with removing the old tiffs from the folder location where are the imagery is stored?
... View more
06-02-2015
04:52 AM
|
1
|
12
|
1585
|
POST
|
For the pour points, I would test Sephe's suggestion first to see what you get but I suspect you may get watersheds for each cell in the pond. I haven't tested this approach so it would be beneficial for you to try it. Even if you get a bunch of small watersheds, you could post process the data into a single watershed for the whole pond. Alternatively, if the pond has an outlet, dam, levee, spillway, etc., you could place a point there or if all land around the pond drains into the pond, you could use the pond's centroid. General knowledge of the slope of the terrain and hydrology for the area will play into where you place a point. As Sephe pointed out, you will need to experiment with the z limit parameter to avoid having your pond filled.
... View more
05-14-2015
04:57 AM
|
1
|
0
|
750
|
POST
|
Thanks for the update. Before we can explore projecting the data to State Plane Feet, there is some additional factors we need to consider: 1. Did you get this data directly from the vendor? The reason I ask is because the vendor may have made a mistake or someone in possession of it after delivery by the vendor may have altered the spatial reference or its definition. The metadata reports it should be in feet so you may just be able to go straight to the vendor and point this out. Alternatively, if you got it from a third-party, you may have to address it yourself. 2. The next question is if the LAS file is projected correctly? What I mean is whether the LAS file was projected to State Plane Meters or if the spatial reference definition was somehow altered accidentally to report that the linear unit is meters (when it is actually feet). If it is the former, you will need to project the LAS file to State Plane Feet and if it is the latter, you need to correct the spatial reference definition. Both should be doable but you first need to identify which scenario is in play. To do this, set the PCS of the LAS Dataset to the same PCS that is reported for the LAS file (State Plane Meters). In a blank map document, add some reference data that you know is spatially accurate for the area of interest. An Esri basemap will do if you don't have anything else you would prefer to use. After adding the reference data, add the LAS Dataset to the map document. If the LAS Dataset shows up in the correct location with the correct extent when compared to the reference data, then you know the LAS file is actually projected in State Plane Meters. If the LAS Dataset is in the wrong location (the misalignment should be considerable between feet and meters), then you know it is the projection definition that is incorrect. Note: If you see a slight misalignment, it could be that you need to specify a transformation under the Coordinate system tab of the Data Frame Properties (or that there just isn't perfect agreement between the LAS file and the reference data). Once you have determined which scenario you are encountering with the LAS file, let me know and we can identify the best approach to take to remedy the issue. Also, let me know which version of Desktop you are using since this will impact what options are available to you.
... View more
05-14-2015
04:38 AM
|
1
|
1
|
1539
|
POST
|
Awesome! Glad it ultimately turned out to be an easy fix. I am not surprised Timothy Hales had the correct solution. I may have learned a thing or two about working with raster data in Desktop from him.
... View more
05-13-2015
08:12 AM
|
1
|
0
|
1301
|
POST
|
I am happy to assist if I can. Mosaic datasets can be pretty complex beasts so this may take some back and forth. I am pondering the possibilities based on the current info but after reviewing the screenshots, they all seem to suspiciously be displaying ~20 datasets. I know you mentioned adjusting the "Maximum number per mosaic" setting but can you verify the setting is actually set to 50 and also that you have removed and re-added the mosaic dataset to ArcMap? This is worth double-checking before we delve into the other possibilities.
... View more
05-13-2015
07:37 AM
|
1
|
0
|
1301
|
POST
|
So just to verify, when you set the Mosaic Method to 'Seamline' and Mosaic Operator to 'Blend,' you are observing the issue? Would it be possible for you to post a screenshot depicting an example of the issue (with the seamlines layer turned on). Also, what Computation Method and Sort Method did you specify when building the seamlines? These are the first places I would begin investigating. If your current solution meets your needs, don't feel compelled to provide this info. I am inquiring just in case you are interested in figuring out what was causing the unexpected behavior.
... View more
05-13-2015
06:06 AM
|
1
|
3
|
1301
|
POST
|
Hi Craig, Looking at the screenshots you provided, the extents reported in both dialogs are very similar. The LAS Dataset is reporting the extent in feet and the LAS File Properties is reporting it in meters. If you convert the extents reported by the LAS Dataset from feet to meters, the resulting values are very close to what is reported in the LAS file's properties. Since we know you have specified the LAS Dataset's PCS to be in Florida State Plane Feet, then my initial guess would be that the PCS of the actual LAS file is really in Florida State Plane Meters (even though the metadata says it should be feet). Have you hit the plus sign beside "Spatial Reference" entry in the LAS File Properties dialog to review the full PCS info? The screenshot only shows the first part of the PCS name.
... View more
05-13-2015
05:41 AM
|
1
|
3
|
1539
|
POST
|
Hi Will, Glad to see you found an approach that works. Just curious, was there any overlap between the source images (or their footprints) or were they rotated tiles?
... View more
05-13-2015
05:01 AM
|
0
|
5
|
3652
|
POST
|
If you have access to 3D Analyst but for some reason not Spatial Analyst, you could use Add Surface Information—Help | ArcGIS for Desktop .
... View more
04-27-2015
05:10 AM
|
1
|
1
|
544
|
POST
|
Hi Ben, I removed the trailing space from the "Latitude" field name and was able to display your points in ArcMap. Give that a shot.
... View more
04-22-2015
04:58 AM
|
0
|
0
|
775
|
POST
|
Hi Kurt, You could accomplish this in a geodatabase using one polygon feature class that contains all grantees. The feature class would need a field that contains the grantee and the polygons could be either the entire area a grantee is responsible for or the individual administrative units. Entire area would be a coarser level of geographic information while individual administrative units would be a finer level of information. In either case, you just need to include the relevant attributes (fields), such as grantee, grantee creation date and administrative unit. You could then just query the feature class whenever you needed specific information and use definition queries in map documents when you just wanted to display specific grantees. Since all the shapefiles have an identical table structure, this should be a straightforward process. If you need guidance on the actual steps to go from multiple shapefiles to a single feature class, just let me know.
... View more
04-22-2015
04:13 AM
|
2
|
0
|
527
|
POST
|
Hi Nancy, It appears to be a datum issue or an improperly defined coordinate system. I tried to reproduce the behavior using the procedure suggested for this discussion but everything worked correctly. A few questions: 1 What is the coordinate system of the data frame (GCS and if used, PCS)? 2. What, if any, transformations are being used for the data frame? If there are multiple GCS's being transformed to the data frame GCS, then list each GCS and its transformation. 3. What is the coordinate system of Export_Output_3? Check the layer's properties to verify it is what you think it is. 4. When you projected Export_Output_3_Project to British National Grid, what geographic transformation did you select?
... View more
04-21-2015
04:57 AM
|
0
|
0
|
3963
|
Title | Kudos | Posted |
---|---|---|
1 | 05-13-2015 05:41 AM | |
1 | 07-22-2015 04:54 AM | |
1 | 07-21-2015 04:58 AM | |
1 | 05-14-2015 04:38 AM | |
1 | 04-16-2015 12:16 PM |
Online Status |
Offline
|
Date Last Visited |
11-11-2020
02:23 AM
|