Latest Contributions by DuncanHornby

Hi Jeff, Thanks for pointing me to the sentinel 2 imagery layer. This is indeed giving me that much needed acquisition date. It's a shame that the wayback data does not seem to keep that, as the quality of those images are very good. May be moving forward, future versions of the wayback metadata could provide such information? I can only but hope! Duncan

Hi Jeff, Thanks for finding time to help me understand the data. So using my example in the original question, am I correct in saying that the way back images were archived in 2019 and that knowing from the glacier dataset I have it was showing a 2003 image when it was archived. But since then the tile layer has been rebuilt and has overwritten (?) the original source image date with 2024 date? Would that mean as a user I can never truly know what the source image date is as the metadata is reporting the built date of the tile layer? Or have I got the wrong end of the stick?

For the metadata the url was - https://metadata.maptiles.arcgis.com/arcgis/services For the imagery the url was - https://wayback.maptiles.arcgis.com/arcgis/rest/services/World_Imagery/MapServer

Just thinking out aloud, when you extended the line have you mistakenly created a multi-part feature? Just go into edit mode and use the vertices tool and in the panel you'll see it list the vertices and if it has grouped them into parts that may be the source of your issue?

I like the idea of it duplicating the last clause but would find it highly irritating if it did it automatically! I would say the opposite to you I don't always add a near identical line to the next clause statement. I think the interface needs to be tweaked and some sort of button on the far right (next to the red cross) that duplicates the last clause would be my preferred solution.

Hi Jon, Thanks for the pointer to the website, I explored this and the info box that pops up provides even less detail than the metadata layer when loaded in ArcPro. So it does not resolve my query. Useful website for quickly checking imagery. Duncan

I'm accessing wayback imagery to update a dataset of glacial fronts on South Georgia. ArcPro and the wayback imagery offer a convenient dataset to digitise against, but ONLY if you know the date of the image. Hence I also access the equivalent wayback metadata for the same release of the imagery. Now I know and expect some imagery I see is actually from a few years earlier and that world imagery has yet to update with new imagery, this make sense and I think something acceptable. It's when it some how reports the future that I am highly suspicious! It undermines the confidence of what I am actually viewing. Can someone from the esri imagery team explain why I am seeing what I see below? I have the 2019 (r00) turned on along with its equivalent 2019_r00 metadata. I have the historic glacial front dataset showing part of a 2003 edge selected. The 2003 data was captured using landsat 7. Note that the 2003 edge aligns wonderfully with the wayback imagery of 2019. Now my small brain is telling me that the world imagery of 2019 had not updated and therefore is showing 2003 images. That would make logical sense, except if you look at the pop-up on the right the metadata is reporting a 2024 image, 5 years into the future?

@Keith-VanGraafeiland thank you for confirming the source of the missing data is NASA and not an error in the processing chain. I still think you should update the metadata for this Authoritative dataset to say that it is incomplete and especially the team behind the aggregate tool to improve error messaging. With regards to the inputs to the aggregate tool, I point you to the comments for the dataset, you'll see I was having difficultly in preparing a subset of the data in crf format. I was not trying to directly process an image service. It is useful for others reading this idea to hammer home the input data type.

Esri, I'm working on a project that requires me to access the chlorophyll-a living atlas dataset here. I am accessing it via the portal tab in the ArcPro catalog pane. As I wish to do some basic mean values over specific months I am using the Analytic renderer which in ArcPro says it is a log10 scale, but this hides the original pixel value. You can see the raw pixel value if you change the pop-up configuration to show it. But these values do not appear to be exposed to the aggregate multi-dimensional tool. SO when I run that tool I aggregate over the log10 values. After asking an AI I have learnt that a raster processing template is applied to transform the values to a log10 scale and the renderer shows this as 0 to 100. No doubt the mathematicians amongst you would say you can reverse engineer the transformation to get to the original chlorophyll-a concentrations but after querying AI it suggests that the common equation used has a gain/offset value which it took a best guess at. I show the response of the AI below but there is no information I can find in the metadata of this Authoritative dataset that describes the exact processing template and the functions used with their settings. So my question is simply what are the raster processing functions applied to the chlorophyll-a data? In what order and what are the settings? I need to reverse engineering this log scale to see the real values for my colleagues analysis to continue. The AI response was this: 2. The Conversion Formula In many oceanographic and satellite products (like those from MODIS or Sentinel), the relationship between the Raw Pixel Value (PV) and the Physical Value (V) follows a specific linear-log transformation. The most common formula used in these analytic renderers is: V = 10^(a . PV + b) Where: PV: The raw service pixel value (0.207422) a: A scaling slope (often 6.75 or 7.0 in standard ocean color products) b: An offset (often -0.9 or -1.0) Let's test the math: If we plug your values into a standard scaling used by many Chlorophyll-A services: Scale the raw value: (6.75 * 0.207422) + 0.05 = 1.4501 Apply the Power of 10: 10^1.4501 = 28.19 Your result of 27.4969 is remarkably close to this. The slight difference comes down to the specific Gain and Offset parameters defined in the metadata of your specific raster service.

‎03-30-2026

I would like to see a new button or menu option that clears all selection on ALL maps in one go. To me it make sense that it becomes an option in the context menu when you click on the maps node in the catalog pane as I show below, but if esri wisdom see fit to place it elsewhere that's fine. Why would I want this? I am using the presentation mode built into arcpro. During rehearsing I click on map(s) and introduce a selection. In a multi-map project, which one would expect if using presentation mode, it become tedious to enter each map to clear the selections on one or more layers so the presentation is left ready to go. It would be good to have a button that instantly clears all selections on layers/tables in all maps.

‎03-27-2026

@BBicking1 - Much appreciated!

‎03-27-2026

I've just wasted hours of my life because this tool does not return a helpful error message. I was trying to aggregate chlorophyll-a data in this living atlas multidimensional datasets. I'm aggregating over user defined time ranges and the tool bombs out returning this error message Failed to execute. Parameters are not valid. ERROR 000800: The value is not a member of 2002-07-04T00:00:00 | 2002-07-05T00:00:00 | 2002-07-06T00:00:00 | 2002-07-07T00:00:00 | 2002-07-08T00:00:00 | 2002-07-09T00:00:00 | 2002-07-10T00:00:00 | 2002-07-11T00:00:00 | 2002-07-12T00:00:00 | 2002-07-13T00:00:00 | 2002-07-14T00:00:00 | 2002-07-15T00:00:00 | 2002-07-16T00:00:00 | 2002-07-17T00:00:00 | 2002-07-18T00:00:00 | 2002-07-19T00:00:00 | 2002-07-20T00:00:00 | 2002-07-21T00:00:00 | 2002-07-22T00:00:00 | 2002-07-23T00:00:00 | 2002-07-24T00:00:00....... Now you might say, well its obvious, one of your parameters is not in the date list. Well I'm being very kind to you by not showing you the 4000 or so other lines in this error message! Every date since July 2002! So the error message says my date is not in list of valid values, but which is it!!!! The tools error message needs to state at the top which input value is missing from the accepted list not just say a value is missing. Multidimensional datasets are very large! I've had to spend several hours to track down that the multidimensional data is actually missing some dates and it was just really bad luck for me that I happen to have chosen one of these missing dates. I hope someone from esri contacts the oceans team and points this out to them, as users need to be made aware that this dataset that is an an Authoritative dataset on living atlas is actually incomplete! For the record I discovered that 2020-06-29 and what was the killer for me 2020-07-01 are missing, there may be more?

‎03-20-2026

@vanesch I clicked on your link out of interest but AGOL blocks it as it appears you are not sharing it.

‎03-20-2026

Yeah I definitely will zoom into your red box limits and you are right I would still like to see an arrow. So I think my "80% position" is actually 80% of the displayed line. So when zoomed out and viewing the whole line 80% is the 80% position, but when zoomed into a junction and lets say you are viewing 50% of the actual line then my 80% position now becomes 80% of 50% of the line...my brain hurts! So the arrow dynamically moves. I think having the user choose a position and default could be mid-point is a good solution as may be you have some sort of symbology that actually places a symbol marker at the mid point of the line. In that case a user might say show the arrow at the 25% position so it does not clash with their symbol?

‎03-20-2026

@DanLee I do a lot of work with river networks, so I dream about nodes and lines! 😁 Reading your comment about a distinguishable symbol designs for nodes, I feel that might be problematic as in rivers you can have nodes where for example two lines flow in, 3 lines flow out and 1 line starts and ends at the same node (i.e. loops) and probably a bunch of other nutty scenarios. So I'm leaning to your idea that the direction is best symbolised with an arrow along its length. Although what you say about it being a mid-point location as that is conceptually easier to understand, I as someone who stares at rivers all day think there may be merit in a to-end position. So below is me using a standard arrow symbol at 50% along length of line. I've made nodes and dangles as solid circles (its what I like). 50% along This is me fudging an 80% along symbol, I just prefer it, although difference in minimal? But having the ability to flash on/off an arrow would be very useful. 80% along

Online Status	Offline
Date Last Visited	2 weeks ago

My Ideas

Latest Contributions by DuncanHornby

Re: Wayback imagery and wayback metadata do not tally, any explanation?

Re: Wayback imagery and wayback metadata do not tally, any explanation?

Re: Wayback imagery and wayback metadata do not tally, any explanation?

Re: Duplicate Symbols on extended line.

Re: Replicate Definition Query Clause when adding a new one

Re: Wayback imagery and wayback metadata do not tally, any explanation?

Wayback imagery and wayback metadata do not tally, any explanation?

Re: Improve error message in Aggregate Multidimensional Raster tool

What are the raster functions applied to the living atlas chlorophyll-a data?

Add the ability to clear all selections on all maps in a project

Re: Improve error message in Aggregate Multidimensional Raster tool

Improve error message in Aggregate Multidimensional Raster tool

Re: Limit or Disable Drag and Drop from Catalog

Re: Add start and end symbology to vertices and nodes display option

Re: Add start and end symbology to vertices and nodes display option

Re: Wayback imagery and wayback metadata do not ta...

Re: Wayback imagery and wayback metadata do not ta...

Wayback imagery and wayback metadata do not tally,...

Add the ability to clear all selections on all map...

Add start and end symbology to vertices and nodes ...

ArcGIS Arcade

Coordinate Reference Systems