Openstreetmap Notes in Ottawa

Posted on by

kicking off the long weekend here in canada in style. friday with some openstreetmap! openstreetmap added a new feature called notes that allows anyone to create an annotated message of a feature/item that needs to be added/fixed etc without needing an account. basically someone with an account attempts to add the note. great way to lower the barrier to reach a larger community that might not want to invest in learning how to use an editor & tagging. i can certainly see this being the case for small business owners that just want to get their location on the map.

anyhow, there’s a great stats page listing contributors here. i’ve wanted to query ottawa & see how many notes there have been.

bbox_ottawa_notes

used to qgis dev (which is amazing! mac dev builds) to import openstreetmap, creating a bounding box polygon & extracting the bbox coordinates. this could definitely be refined. in a nut shell you pass the bbox to the osm notes api & it returns xml (default) or several other formats. the start represents the only note in the region in gpx.

api call:

curl http://api.openstreetmap.org/api/0.6/notes?bbox=-76.30,45.09,-74.96,45.62?closed=-1

single note details — red star in graphic:

<?xml version=”1.0″ encoding=”UTF-8″?> <osm version=”0.6″ generator=”OpenStreetMap server”> <note lon=”-75.6967664″ lat=”45.421061″> <id>3125</id> <url>http://api.openstreetmap.org/api/0.6/notes/3125</url&gt; <comment_url>http://api.openstreetmap.org/api/0.6/notes/3125/comment</comment_url&gt; <close_url>http://api.openstreetmap.org/api/0.6/notes/3125/close</close_url&gt; <date_created>2013-05-14 22:33:23 UTC</date_created> <status>open</status> <comments> <comment> <date>2013-05-14 22:33:24 UTC</date> <uid>1679</uid> <user>andrewpmk</user> <user_url>http://api.openstreetmap.org/user/andrewpmk</user_url&gt; <text>Ottawa light rail, construction will begin soon. Also missing major bus routes on transitway like #95.</text> <html>&lt;p&gt;Ottawa light rail, construction will begin soon. Also missing major bus routes on transitway like #95.&lt;/p&gt;</html> </comment> </comments> </note> </osm>

really need to prompt the notes features in my neck of the woods. it’s a great feature. next steps would be to take this into python & clean it up, gives me a good excuse to keep learning python!

enjoy,

richard

Random Assortment of Todays Findings

Posted on by

SAGA — Version 2.0.8 on the saga_cmd side for api access. Overland Flow Distance To Channel Network does not work. Had to upgrade saga to latest beta version & then not issues. You can use that module to calculate catchment gradients etc, pass in a rasterized lake for channel network & then:

gradient = vertical/overland

length gradient = overland/gradient

 

Hillshades in FME 2013 sp1 — to get a vertical exaggeration, add a RasterSingularCellValueCalculator before the HillShader.  Set operation to “*” & value to exaggeration value. Not sure why Safe didn’t include a vertical exaggeration in HillShader, hope they do.

 

Enjoy,

Richard

Installing QGIS Master (1.9) on Mac OSX

Posted on by

just a few links that made the process of installing qgis master aka 1.9 development (soon to be 2.0; sometime this summer) on a mac osx. i’m running lion 10.7 with mac book pro. honestly, these folks do such an amazing job & free to provide these frameworks & builds; so kudos!!!

from qgis page, you can install either current qgis 1.8 or dev 1.9.  regardless you will need the frameworks provided by KyngChaos, install the GDAL & GSL complete; add GRASS if needed. i installed the nightly dev (1.9) builds provided by larry shaffer of dakota cartography, you will need to install GDAL 1.10 complete in order for the build to fire up on your box.

seriously, it’s that easy. install the frameworks & download nightly zip & unpack & double click the dmg. to start qgis from command line, you will need to add an alias, i’ll let you figure that out:)

 

enjoy,

richard

 

newest free semi-national 1 arc sec srtm derived elevation data -> cdsm

Posted on by

don’t like integer cded? probably not alone in that regard but there’s a partial answer. at least if you need data below 60 deg north. enter the  canadian digital surface model (cdsm), produced by the folks out at centre for topographic information (cits) in sherbrooke quebec. it’s floating point data to start with & is the best publicly available elevation data in canada. i’m currently using it for lake basin delineation & have been very happy with the results.

technical details below:

0.75 Second Canadian Digital Surface Model (CDSM) Version 1.0:

The national 0.75-second (~20 m) CDSM consists of a derived product from the original 1-second (30 m) Shuttle Radar Topographic Mission (SRTM) digital surface model (DSM). The SRTM data were reprocessed from their original form as follows: 1) the data was void-filled using the Canadian Digital Elevation Model (CDEM) with the Delta Fill Surface method from Grohman et al. (2006); 2) the Vertical Datum was changed from EGM96 to CGVD28; 3) the data was projected to UTM using cubic convolution; 4) the data was smoothed using the Denoise algorithm from Sun et al. (2010); 5) the data was re-projected from WGS84 to NAD83 and aligned to the 0.75-second CDEM grid resolution using cubic convolution; and, 6) the waterbodies for which new elevations were calculated due to the change of vertical datum were reflattened.

you can download it from the new geogratis portal., which by the way is a massive improvement over the older version. if you need it in larger swaths or by nts index, use the ftp site for download.

colored hillshaded cdsm 30m created in saga gis to give you an idea of what the data set looks like.

cdsm_30m_hillshade

enjoy,

richard

my webinar rpm presentation details for 1330hr est

Posted on by

giving a webinar based remote predictive mapping (rpm) talk today. details below.

 

RPM Technical Transfer Presentation Series 2012- 2013
A Collaborative Series on RPM Methods
Geological Survey of Canada and Agriculture Canada
Final lecture of the season

Friday, May 3rd @ 1330hr. 

WebEx details below.

Richard Burcher
Combining Open Source software and public geodata with
Remote Predictive Mapping (RPM).
Regional scale detection, characterization and extraction of streamline features on
Eastern Victoria Island.

Abstract
This talk will present research that utilized Open Source software and publicly available regional geodatasets in the development of successful and adaptable RPM work flows. The work flows  detect, characterize and extract streamline features objectively across the landscape of Eastern Victoria Island. The treeless, Arctic region readily displays a range of shapes and sizes of streamlined forms, most are visually discernible as small to medium-length, spectrally-bright, elongated ridge tops. Streamlined features were defined as elongated, narrow objects that carry well-defined ridge tops. This definition was expanded for spectral data to include highly reflective linearly features.

A wide range of Open Source software was used to develop the work flows within the Windows operating system. Both pixel and Object Oriented Image Analysis (OBIA) based approaches were implemented. Key advanced computing techniques included: Kmeans, Maximum Likelihood and random-Forest machine-learning algorithms, morphological image processing, morphological spatial pattern analysis and multi-resolution image segmentation. Predicted streamlined features were assigned a series of descriptive metrics, that include a calculated azimuth from a custom shape-metrics toolkit.

Several publicly available regional scale data sets were evaluated that included: medium resolution GeoBase integer CDED, medium resolution NRCan orthorectified Landsat 7 mosaic
of the Canadian Arctic archipelago and high resolution GeoBase orthorectified SPOT panchromatic. Manually digitized streamlined data (Stokes, 2003) were used as a first order comparison of the predicted RPM streamlined forms. Contour plots of the landform azimuth data portrayed the regional-scale flow domains.

Open Source software and open data provides equal access, collaboration and research opportunities to all, a right of use that is not available in proprietary software. OBIA paired with publicly available regional remotely sensed imagery provides increased spatial resolution and additional spectral information not offered by current elevation data. These tools and data sets provide new research development opportunities for landforms in glaciated terrain by enabling objective, rapid detection, characterization and extraction of glacial landforms, specifically streamlined features.
Large areas can be processed within hours, compared to the manual interpretation project time lines of months to years.

 

WebEx Details Below
——————–

To join the online meeting
Starting time: 1:15 pm, Eastern Standard Time (New York, GMT-05:00)
——————————————————-
1. Go to https://pwgsc-nh.webex.com/pwgsc-nh/j.php?ED=187574582&UID=0&PW=NY2M2MmYxMzcx&RT=NCMxMQ%3D%3D
2. Enter your name and email address.
3. Enter the meeting password: RPM
4. Click “Join Now”.
——————————————————-
To join the teleconference only
——————————————————-
Call-in toll-free number: 1-8774134785 (US)
Call-in number: 1-6139607512 (US)
Attendee access code: 946 926 5
—————————————————–
For assistance
—————————————————–
1. Go to https://pwgsc-nh.webex.com
2. On the left navigation bar, click “Support”.
3. Call 1-800-226-6338 or 613-941-9554

fme — git your workspaces & save the pain of multiple copies. easy rollback if you break it!

Posted on by

quick tip:

use git to version control your fme workspaces. dead easy & saves you the hassle of multiple copies of files & what happens if it worked before you changed something, how do you fix? you do know what you changed right? software development wouldn’t be with source control right, so simplify your life.

on windows, download tortoisegit, install. i’m using a local repo but you could host this say on git hub or elsewhere, fme workbenches are small. so add the workbench, ignore the log files.  work on work bench, any significant changes be sure to commit the changes. it’s so easy!

enjoy,

richard

 

fme 2013 so far — python focused

Posted on by

working with fme desktop 2013 sp1 for the past few couple of weeks for a clients research project. first off, the more time i spend with it, the more it’s value is apparent. yes there is certainly a learning curve but well worth it.

my work involves calculating lake basin watersheds for an entire lake feature (on-demand from user & at national scale), not a single pour point. how do that, download saga & look under the terrain analysis – hydrology – upslope area, the target area attribute is the input rasterized lake.  saga offers an api using saga_cmd to access functions, so how to preprocess lakes ie buffer, generalize, reproject or any number of modifications & interface that newly lake feature to create basins? python is well intergrated in fme, i tried the pythoncaller but have not figured out how to write out single feature as temp shapefile to be read in by saga; had issues with the universalwriter & have filed case. will update when they get back in touch because the ability to write out a temp shapefile or other mid workflow stream, process with someothe software & reread back into the workflow stream is huge. so my approach has been to use scripted python parameters to setup the initial directory structure & covert the elevation geotiff to saga sgrd format. waterbodies get read in, processed & written out as individual lakes where a python shutdown script reads the lakes directory & calculates the lake basin for each file.

i’m sure there are better ways to accomplish this but i’ve only been digging into the python aspect for the last few days. it offers a lot of flexibility!

posted the initial python code for fme shutdown script to interface with saga. hope this gets you going in fme or you can use the example outside of fme, your choice:)

 

enjoy,

richard