Newsflash
| Quantum GIS Version 0.11.0 has been released. It is available in source form, and as binary executables for Microsoft Windows, Mac OS X, and GNU/Linux. All versions can be obtained from our download page. If the version for your platform is not available please check back in a day or two as some packages are still being built. |
| Interview with Brent Wood |
|
For this week's interview, we travel to New Zealand to chat to Brent Wood (known as baw on IRC) to find out how Open Source GIS tools are being applied in the fisheries industry. Brent is co-author of a chapter in Geographic Information Systems in Fisheries and has presented papers on the use of Open Source software in the fishery industry at conferences held by the Fishery-Aquatic Reseach Group. The interview was conducted by Tim Sutton, Tyler Mitchell and Gary Sherman over IRC.
TS: Brent, welcome to this the 7th QGISSER interview! Could you give our gentle readers a brief biography of yourself? What is your name, where do you live and how many cats/dogs/kids etc do you have?
TM: I understand that you spend a lot of time out on the ocean. What are you up to when you are out there and do you use any open source mapping tools while your are out there? My main job is a pretty varied role in fisheries research. Biological, GIS, IT support. Pretty much any fisheries data has an important spatial component, so GIS is a pretty standard tool. Producing various maps for planning & publication, navigational stuff. Also now using QGIS with navigational data & other spatial data while I'm out there (& hopefully increasing this aspect) TS: So could you give us an example of what kind of analyses you do? Did you once send us a sceenie of trawl transects and catch per unit effort calculations done with the aid of GIS? How do you assemble such an analysis - I mean what are the software building blocks you use? Depends on the type of data, if it's a trawl based survey, then we have a random stratified survey design, tow data is entered from a paper log kept on the bridge, catch info is captured directly using PC's & digitisers with custom software. It all finishes up in an RDBMS. Most of the analysis is done with R or custom software, the maps I generally use GMT for. For acoustic surveys, Gavin [Macaulay] has set up the acoustic database in Postgres, then I showed him QGIS & Postgis & he has jumped right on board. I have a script which parses the QGIS project file to generate publication quality maps using GMT.
TS: How long have you been a QGIS user and how would you characterise your useage of QGIS? Casual user, power user, sometime developer, serious developer? I think I started with v0.4. Played with GRASS, but until recently it was pretty limited in vector capability. I used the commercial GIS Genamap, fantastic topological vector GIS, but I wanted an OS tool. I have looked at Thuban, hoped FMAPS would turn into something useful, JUMP is in the mix, then found QGIS & a great team of helpful developers. I'd say I'm a user, possibly power user. My only development skills amount to an aptitude for screwing things up which helps find bugs.
TS: What is your employers attitude towards the use of OS GIS? Are they 'any tool that does the job' folks, or are you using Open Source tools by stealth introduction? The institute I work at is largely an ESRI shop GIS-wise, but as a research organisation working in meteorology, oceanography & fisheries there is a very wide range of software used. R is probably the predominant stats tool, but if you include stats & database tools used for working with spatial data & generating tables/plots/maps, there are probably 20 or more packages used. There is plenty of Linux, a bit of NetBSD, no policy against Open Source. Individual researchers are reasonably free to choose what tools they want, so there is a fair bit of interest in OS tools at grass roots level. TS: And does the choice of OS over commercial software tools usually center around cost? Or are there other incentives for you to be using OS tools? Cost is always an issue, but compatability & suitability are probably pre-eminent. I got into OS because I could not afford commercial tools. Prior to my fisheries GIS role I was more into data/database admin, so much of my GIS focus & interest is in spatial data management, so of course PostGIS is way up there for me. Also data visualisation, which GIS does OK in 2D but for midwater fisheries work it is a bit more limited. I've done some work with OpenDX for 3D data visualisation. TS: If you had a bounty list of 5 most wanted features for QGIS, what would they be?
TS: How would 1) work? I mean how would we determine the correct order in which to connect the points to form verticies of a line? Natural order of returned query? And similarly, how would we determine which points belong together to form a line? baw: well, I'd envisage it something for building a track (line) from individually logged GPS points, so sequentiality is explicit, by timestamp, line-in-file, etc. One frustration with PostGIS after Genamap is it's limitations in this area. For example: "select boundary(memgeomunion(geometry))" returns a line (or multiline). But it is far from trivial to get it into a polygon for further point-in-poly queries. I'm not sure it is really a QGIS issue, maybe more appropriate for GEOS or PostGIS. JUMP seems to be the most effective tool for this, but it is the one area of GIS where I have found the OS tools to be inferior to some commercial GIS tools. TS: Is it true that there are plenty of fish in the sea? Can you give us a definitive count of just how many are left? Nope. Not definitive. One thing that the international history of fishing tells us is that its a safe bet that there are less than last year. Maybe not all fisheries, but certainly many. GS: What would you like to see in terms of QGIS being able to integrate with other tools? I think you've done a brilliant job in this area. QGIS is interoperable with most of the mainstream OS tools out there, at least those I'm familiar with. I'd like some other tools (like GMT) to become more interoperable, but I think that requires the developers of such packages to see such integration as a priority & work towards it. Frankly, I wouldn't change a thing about how you are going with QGIS. It's effective, consultative, and I get to do things I like with the results of your labour! TM: You have been involved somehow with compiling datasets for New Zealand. Can you tell us a little more about your involvement, interest and hopes for how country-wide datasets will get used? Ouch... I guess it goes back to my database origins. I saw all the neat tools you & the rest of the OS GIS community are providing & figured a nice, local dataset was needed to really make them useful here. The NZ Govt has a pretty good dataset, which is supposedly "free" but in reality worked out at least $2000. Genamap donated a copy of the data and I found a co-conspirator to do the reformatting of the data into a non-proprietary format (shapefile). So now NZ has a freely distributable very complete set of 1:50,000 vector topo data to go along with your free software. 
TS: Well I'd like to thank you very much for taking the time to be our 'QGISSER of the week'! Many thanks for all the useful contributions |
Brent Wood; age 46; married with a 14 yr old daughter, around the house we have about 5 hens, 3 cats, 2 goats & a couple of guinea pigs. [I was] born & am still residing in New Zealand, just out of Wellington.
