Global Mapper v25.0

new function: water level rise ...

gm_user1
gm_user1 Global Mapper UserTrusted User
edited February 2013 in Technical Support
By playing with the new analysis function for "water level rise", some aspects puzzle me and I need some clarifications:

1 - We use an input data (90m-res SRTM) and the default setting for the above analysis (see the figure below):

GUI_options.jpgdemo.jpg

The pink area shows the analysed result below 5m threshold. But why some pixels with heights 1m or 2m are not being picked up? Is the approach based on a 'bathtub" concept?

Some pixels nearby have elevations less than zero meters. If I set a smaller number say "-10" for the "Elevation (i.e. Sea Level", the tool run prompts a message window saying "Unable to find any streams of sufficient flow in designated area".

2 - The DEM data is manually set as not resampled; i.e. use NN resampling. If I analyse the entire data, the picked-up area fits the pixel square well (top picture), which is expected. But If I analyse the DEM AOI (seen on the screen part only), the analysed result is shifted and it seems this is still automatically based on a resampled DEM.

Thanks.

Comments

  • global_mapper
    global_mapper Administrator
    edited February 2013
    The water rise does indeed use a "bathtub" model, so far a sample location to be added to the result water dropped on that sample must flow to a target cell. With the options that you selected to flow to a single height of 0 meters only those cells that flow to a cell with an elevation of exactly 0 meters (not below) will be used.

    For SRTM with some noisy negative samples you might want to replace the negative values with 0 (use the Alter Elevation tab of the Options dialog), or a better approach is to use area features for the ocean and select to find the rise height from the selected areas. This is a more powerful option as you are not limited to a single height and can model the rise of features with a non-constant Z, like flood plain areas.

    I did decide that the behavior might be better if it allowed going from <= the specified elevation and not exactly what you specified, so you'll find flow to anything at or below 0. Keep in mind you would also get areas in inland low-lying areas, like Death Valley, but those areas would be disconnected and easily identified. I have placed a new build at http://www.bluemarblegeo.com/downloads/global-mapper/global_mapper14.zip with the latest changes for you to try. Simply download that file and extract the contents into your existing v14.xx installation folder to give it a try. If you are using the 64-bit v14 version there is a new build at http://www.bluemarblegeo.com/downloads/global-mapper/global_mapper14_64bit.zip .

    Thanks,

    Mike
    Global Mapper Guru
    gmsupport@bluemarblegeo.com
    http://www.bluemarblegeo.com/
  • gm_user1
    gm_user1 Global Mapper User Trusted User
    edited February 2013
    Thanks for the insight and the fast response. We all know you are providing the first-rate support.

    1. I use the SRTM data with the water level of zero metres and the default setting for "water level increase amount" of 5m. The figure below shows the pick-up result in pink. But why are those pixels with yellow boxes (elevation <= 5m still) not being classified? I wonder you may have referenced some river network structure information, so I try to superimpose creeks (blue lines) and watershed boundaries (white dash line). Please confirm.

    2 - I feel it is better to "enforce" the analysis area spanning from exact pixel boundaries / corners, regardless to analyse either the entire data or an AOI on the screen, so that the derived result (water level rise areas) would not contain "analysis uncertainty". See the bottom figure in my initial post, where the derived result is shifted if I analyse an AOI on the screen, not in line with the exact - physical meaningful - pixel boundary any more. This issue might be generic and applicable to other similar analysis functions.

    3 - Regarding the non-constant Z increase over a polygon area (e.g. flood plain): Let's say we have a large flood plain with an elevation of 1m in the river mouth and 10m in the upper stream; elevation monotonically increases from river mouth to upper stream. If we set "water level increase amount" of 5m, does that mean all pixels that are adjacent to the known flood plain area (polygon) and with an elevation less than 15m (a max 10m + 5m) should be picked up? Probably not, then how do you increase 5m non-uniformly (i.e. 6m for the river mouth and 15m for the upper stream)? Any boundary conditions or thresholds, or some fluid processes involved?

    Thanks.

    water_level_demo2.jpg
  • global_mapper
    global_mapper Administrator
    edited February 2013
    I just went ahead and added the standard Resampling method selector to the water rise and watershed dialogs so you can choose exactly what resampling that you want. If you also want the cells to perfectly align with the originals, check the option in the Advanced section of the General tab of the Configuration dialog for exports to snap the bounds to pixel boundaries if possible. I have placed a new build at http://www.bluemarblegeo.com/downloads/global-mapper/global_mapper14.zip with the latest changes for you to try. Simply download that file and extract the contents into your existing v14.xx installation folder to give it a try. If you are using the 64-bit v14 version there is a new build at http://www.bluemarblegeo.com/downloads/global-mapper/global_mapper14_64bit.zip .

    The water rise uses the same flow analysis method as the watershed stuff. So every cell gets a D8 flow direction. For both a fixed height and an "area rise" calculation, the flow path from each cell is traced until it hits a cell that is in the "destination" (i.e. <= the base height or in a flow area). The difference in elevation between the source cell and the first encountered cell in the destination area is calculated and if <= the rise height then that cell is included in the new area. So for a flood plain if you are upstream and a cell at 8m elevation hits the flood plain at a cell that is 5m in height, the difference is 3m so it would be included in a water rise of 3m or more.

    For your cells that weren't hit, did you specify a depression fill depth when doing the water rise calculation? If your cells were in a depression and didn't have a flow path to reach the ocean then they wouldn't be included. However you would also want your depression fill depth to be at least the rise height. Actually I should probably just enforce that directly, allowing you to go above but not below the rise height in the depression fill depth. (The build above now has this enforced).

    Thanks,

    Mike
    Global Mapper Guru
    gmsupport@bluemarblegeo.com
    http://www.bluemarblegeo.com/
  • gm_user1
    gm_user1 Global Mapper User Trusted User
    edited February 2013
    Thanks for the detailed information.

    The cell (from the figure above) with "4" is indeed a depression. But after running the new build with the reinforced fill of depression, the result is still the same - not picking up. I can observe the value is reset to "5" after filling, which is correct and expected. The issue might be related to the case below - dealing with the exact value of 5m set in the max amount for testing.

    I am still trying to figure out why those cells with "5" (marked with yellow box) are not picked up; they are located in the right flow direction. It is likely to do with the rigid value set for the max amount (i.e. 5m)?
  • global_mapper
    global_mapper Administrator
    edited February 2013
    Can you provide the data file that you are working with so I can test this particular area and see? Also if the value is 5 and your rise was set to 5 maybe that is the issue. If you set the rise to say 6 does that grab the cells?

    Thanks,

    Mike
    Global Mapper Guru
    gmsupport@bluemarblegeo.com
    http://www.bluemarblegeo.com/
  • global_mapper
    global_mapper Administrator
    edited February 2013
    I found an additional issue caused by my recent change to allow going under the specified single value. The max difference in elevation was checked relative to the first sample in the area so if you had values below zero the difference had to be relative to that. So a height of 5 meters would only go up to 4m if the closest point was -1. This is how it should work for areas, like flood plains, but not single fixed elevation. I have placed a new build at http://www.bluemarblegeo.com/downloads/global-mapper/global_mapper14.zip with the latest changes for you to try. Simply download that file and extract the contents into your existing v14.xx installation folder to give it a try. If you are using the 64-bit v14 version there is a new build at http://www.bluemarblegeo.com/downloads/global-mapper/global_mapper14_64bit.zip .

    I also made the advanced option to snap the bounds to the pixel bounds of the layers work for the watershed/rise calculations as it wasn't applying there.

    Thanks,

    Mike
    Global Mapper Guru
    gmsupport@bluemarblegeo.com
    http://www.bluemarblegeo.com/
  • gm_user1
    gm_user1 Global Mapper User Trusted User
    edited February 2013
    This has solved the issues so far.

    1 - For this test data set, if we set the max amount of 5.001m, instead of 5m, the hit area meets expectation exactly. (I can verify this with the "bathtub" result from another tool.) Interestingly, if we set 6m, all pixels <=6 can be picked up correctly. It seems this is only to do with the exact threshold of 5m.

    2 - You said you have added an "advanced option to snap the bounds to the pixel bounds ...". I could not find that in the new built under Configure | General | Advanced Options section.

    3 - Sample file is attached. TestData1.zip

    Thanks again.
  • global_mapper
    global_mapper Administrator
    edited February 2013
    The option on the Configuration dialog is about 2/3rds of the way down and labelled 'Export: Snap Export Bounds to Pixel Boundary if Possible'. That should allow everything to line up with the original samples when cropping.

    It sounds like everything else is ok with the new build, correct?

    Thanks,

    Mike
    Global Mapper Guru
    gmsupport@bluemarblegeo.com
    http://www.bluemarblegeo.com/
  • gm_user1
    gm_user1 Global Mapper User Trusted User
    edited February 2013
    From my current exploration, so far so good. Thanks.