A geological map workflow in QGIS with Chris Lambert

alrighty um welcome everyone to our um last session of the qj's open day plugged in um for this session we have chris who's going to be walking us through um a geological map workflow in qgios which sounds fantastic so i'll hand over to chris and he will be um walking you through that chris okay thank you very much i'm just going to share my screen and i'll be with you okay um can anybody hear me yes perfectly we're good and you can see the screen [Music] yeah okay great all right um good afternoon everybody uh thanks very much for having me on this this is um yeah hopefully you'll you'll enjoy this talk uh my name is chris lambert i'm a exploration geologist with srk exploration services based in cardiff in the uk today i'd like to share a workflow in qgis that i've been using recently trying to develop a little bit for digitising geological maps i hope this talk will give you an oversight for the workflow and if you haven't tried it before maybe introduce you to a few of the useful features in queue that i've come across uh it's still a work in progress so if you've got any questions or comments please just chuck them into the chat because i'd be really interested to hear your input um and to be honest there's sort of like new versions of qgis are coming out uh i've been finding that the process is changing uh you know almost kind of monthly at the moment so yeah i'd be interested to hear if there's any new features out there uh okay so i've been using gis software for probably about 10 years now something like that but i've almost entirely converted over to qgis and i've been using q for uh probably about two or three years something something along those lines uh we find it just really valuable when dealing with junior exploration companies who don't have like big budgets for gis packages it's just really helpful to be able to communicate them directly in the software that they're using and things like that so i thought i'd jump in and like quite broad um sorry if there's lots of geologists in the in the audience but i thought i'd just sort of like create a level playing field so everybody kind of is talking off the same page um geological maps are a graphical representation of the distribution and geometry of different geological features below our feet um these include different rocks soils geological structures and other features that we uh that we represent by different colored points polygons and polylines on our maps most of these maps are built from observation data points so it's often um you know geologists going to the field and observing the rocks at a particular outcrop you know collecting multiple uh data points and then extrapolating that information to try and predict the trends of the geological features in the subsurface you know below the sort of the cover um geological maps are used in a wide range of uh uh transsolving and solve a wide range of problems as well in one of the earth resources risk management and engineering sectors uh we use them for land use planning risk mitigation flood analysis agricultural land selection sinkhole monitoring you name it you know they've got you know as long as it's kind of earth related they've got a pretty good use for um but what i want to talk to you more about today is um what i use geological maps for and that's for trying to discover new mineral deposits um it's pretty central to like every my everyday job so yeah that's what i'd like to present you a bit of a workflow on how we try and build these maps i've put these two maps up in particular on our screen um hopefully you can see these okay but uh yeah i kind of just firstly i wanted to show you what a geological map looks like if you haven't seen one before they're quite pretty actually you know they're really colorful nice nice things to work with to be honest and you can be quite artistic when you're working with them as well the one on the left is a map published by william smith in 1815 and this is kind of like well william smith is the probably classic was like the forefather of modern geological mapping really uh it's probably one of the first like uh sort of modern style geological maps to be published and on the the right hand side we've got the um the latest iteration of the the geological map of the uk published by the british geological survey and kind of what i wanted to um really what i wanted to sort of highlight with this slide is that although the technology has changed over the years the principles of mapping for geology and lots of other mapping styles as well have remained quite consistent i thought i'd include this slide as well um because i just wanted to show that although the mapping techniques are really well constrained mapping has been a solution for geological problems for millennia this is the turin map which shows a portion of the egyptian desert and this map dates back from about 1 150 years bc it's it's bound to be one of the oldest geological maps in existence and funnily enough it kind of does exactly what we try to do as geologists today uh it displays different lithologies or like rock types so we've got things like the pink granites we've got dark sandstones uh so pentanites uh it shows logistical aspects as well so we've got you know uh main uh roadways and roots coming into the area we've got the location of the the mining village uh the location of the mine accesses and as well as the sort of the water wells but it also displays uh quartz vein locations and um in particular it sort of highlights areas that are a good gold bearing area so quartz tones that you can work to extract gold um i just think that this map's really important because it shows that even though these processes that we're going to talk about today perhaps you know potentially 3000 years old it's still still considered a key exploration tool today i also thought it would be quite a nice segue into the bulk of my presentation because uh the case study that i want to present you today is also a portion of the egyptian desert and it's also a gold project targeting the same type of quartz veins that ancient miners were looking for as well now due to confidentiality reasons i've had to remove most of the geographical indicators from this talk but it doesn't really matter that much because what i want to show you is the kind of features i've been using in qgis to make my you know these kind of modern geological maps so um yeah the first the first real step in in any of our projects is is locating the project first of all and sort of getting a feel for the area we would typically do this with um freely available base map data and this is um well we we access this in qgis through the the quick map service plugin uh you can download this from the the plugin manager and it gives you access to sort of the esri the google earth the bing satellite imagery the freely available imagery that is just really helpful for you know good overviews of a project what we can observe from this particular area which is what we're going to have a look at today obviously we've kind of got this lake on the far west of the uh of the area we've got a series of river channels running across it but we've also got this kind of east-west striping this um it's the the main lithological the main sort of uh geological strata uh that we're particularly interested in mapping out today in addition to that we've got a series of circular features as well which are granitic intrusions they're granites which have upwelled from deep in the crust and they've come up to surface so we've got a quite a bit going on but hopefully this workflow you know i'm going to demonstrate what i did with this and how to produce a geological map out of it probably one of the you know the key things that you can take from this as well is uh is the incredible outcrop exposure that we can see in the satellite imagery um you know being that it's a you know a really arid environment there's very little vegetation and soil development um it this these kind of areas really lend themselves to satellite interpretation but we can take this a step further and i know some of the talks today have kind of touched on some of the satellite imagery interpretation um and this is just one example of multi-spectral data that we use well yeah we use quite uh quite systematically in my work to be honest uh this one in particular is uh is produced from sentinel-2 satellite imagery but we also use asta quite a lot as well and these these satellite missions allow us to produce false color imagery and undertake mineral mapping and lithological discrimination processes so what that means is effectively we can start to distinguish one rock type from another in this map in particular you know we're looking at you know the colors automatically show us that we've got different you know different things going on in this area so we we're the plan is to start to try and pick these apart and start to really produce a map that um that displays this sort of the interpretation of what's going on this map in particular was completed through srk's internal methodologies but i know for a fact that similar processes can also be done in qgis using the srtn downloader and the semiautomatic classification plug-in uh these are both uh accessible through the plug-in manager and so if you're interested in this i i really recommend that you go and download these plugins um there's tons of tutorials on youtube and in various different formats so you know you once you put some sort of uh put some effort into looking around you can really start to sort of you can you can pull out a lot of this you know a lot of information from this data okay so the first sort of real step in digitizing a map that we'd normally go go through is um undertaking a structural interpretation so really the sort of what we're looking to do here is to try and map out the main geological breaks the discontinuities the fractures the faults that are kind of bounding different packages different domains of rock and we're starting to try and pull this map apart to try and uh yeah just try and work out what's going on in here uh it's important for two main reasons firstly it provides us with a framework for the map so that's you know that's helpful that's what we're gonna hang the rest of our map on so we're gonna build it off this structural interpretation um but also our our expiration model for for gold uh indicates that what the deposits we're looking for are going to be situated in close proximity to some of these major structures so the identification and the categorization of these structures is key to our exploration uh efforts and our methodologies um in terms of the actual digitization of these lines and in terms of the rest of the workflow actually what's really important is um is just being neat and tidy and really systematic with this um this means that we really have to try and ensure that all of the lines are connected and snapped to nodes but also that there aren't any small gaps or you know um dangling kind of overlaps and things like that we've really got to try and make this line work as neat as possible a few tools i found in qgis life um well yeah that i use pretty routinely while going through this stage is the v clean and fixed geometries tool you can access these through the processing toolbox so from your um from qgis from the top window you uh if you go into processing and toolboxes there's a huge raft of different um algorithms and tools that are really helpful for pretty much everything that you need to do um yes you can pretty much just go in and you can start to find some of these tools to help you really like keep clean line work and tidying things up so i use those pretty much yeah systematically as i'm going through this process uh in terms of like actual geological digitization i've got uh probably one tip that i could give would be to um to digitize two polyline layers so why can i divide my digitizing into two two layers uh the first one is my framework layer and that that's pretty much what you can see on the screen here so uh within that i'm digitizing my bounding box so the black line that sort of um that's marking out the main area of the map um i mark i i digitize my contacts so um the kind of where to rock units meet the contact between those two rock units i'm marking out and those are marked in this map in white uh i also mark out a series of unconformities if they're there uh faults thrusts and shears other sort of geological features which are like controlling the geometry of the geology in this um in this example the second polyline layer that i produce is what are terms like an overlay layer so they're features that are um that they're features that are perhaps things that we'd want to present in our final map but aren't fundamental to the uh to the framework of the of the map itself so things that i would include in that are uh general lineations if i want to sort of give the general feel of of how units are sort of um you know kind of displayed across the map i'd mark out the lineations i'd mark out the series of dikes that i can see minor structural features fold trends and other things like that and this like i said this will be displayed on on our final map map but as an overlay layer rather than anything that we're going to work with moving forward from this point uh the second step is quite simple actually um we simply we take our our framework polyline layer and we run it through the polygonize tool again this is accessible through the um the processing toolbox and all it does is it just it fills all of the voids within our um uh within our framework with polygons so it's a really quick and simple method for digitizing the entire map effectively um but it's also why neat line work is so important in this um in this process so if you have um if you have any gaps within your line work your polygons are likely to spill across your contacts and into you know and join up with each other so it's really important as i said before just to keep that line work really neat and tidy and ensure that everything snapped correctly the beauty of qgis though is that whenever you run these tools it creates a temporary layer so it's never destructive it won't corrupt your original file and if there are any mistakes you can simply remove what's just been produced this temporary file go back to your line work fix any errors and then re-run the process so it's um yeah it's a nice iterative process it's really um it's really easy you can sort of like jump between these stages and just really ensure that you you're uh yeah you're nailing it at every stage really okay the third step is a bit more uh i guess it might take a little bit more explaining so i'd be really interested to hear if anybody else is using this kind of methodology um and this step it involves uh producing what what i terms a point seed layer so effectively what we're trying to do is every polygon that has been produced will have a point attached to it a point within its within its geometry um the reason why we do this is because we're going to attribute this these points in a minute and then we can join that with the polygons that we've just created now this can be done in a couple of different ways the first way is to do it manually which is quite slow but it's kind of how i started out this workflow every time or as i was digitizing my um my line framework i was just dropping points into each of those areas and attributing them with whatever information i was trying to collect so in this case it would be uh you know lithotype lithology or rock type or something like that um obviously as you start to build up the size of the maps you're making you know if you're really covering large areas really complex areas doing this manually is quite um quite a job so what i've started to do is take my polygon layer that we produced in the previous slide and i start to run a either a centroid or a random point from in polygon uh algorithm again these can all be found in the processing toolbox and simply all this does is it just it processes through the map quickly and just sticks a single point in every single polygon that you've produced in the previous step so it's a really quick way for generating all of your points um now as i said we're going to have to try and manually sort of attribute this all of these points so again it's quite a sort of labor intensive process i guess it would probably take me about an hour or two to go through this map and sort of attribute all of these points correctly apply any metadata to them but what i've found is there's a couple of ways to speed it up you know there's multiple different select tools in qgis so you can use the lasso tools and other kind of ways to select multiple points that are going to have the same attribute information you know you can move along trends or across certain lithological packages to try and select all of the points that are in the same units and then apply um you know the attributes to them in addition to that i've really started to use the attribute forms as well so if you again if you haven't come across this then i would really recommend going to have a look at these features uh you can access this in the layer properties uh you go down to the attribute forms as being displayed here and then you can there's all these different widget types that you can start to play around with and these widgets just they make uh sort of adding and um verifying the data that you're putting into you know the background of these files it really speeds it up and really gives you some good validation tools to use so i really recommend looking at those um for this process i use value maps quite a lot and what that does is it gives me drop down menus so effectively what i can do is i can fill out the type of rocks that i'm expecting to see on this map uh ahead of time i can apply that and i can go through my map and i can just select multiple points that are going to have that rock type and then the drop down list allows me to enter those those attributes really quickly so yeah it takes a little bit of um a little bit of work but it's it is quite a quick method uh one thing that i have been looking at and i'm wondering whether anyone could help me with this actually is i wonder if there's a something that i can run that will how can i put this um that can categorize the points the point seed layer by the the value of the raster on which they're sitting on so um effectively what that will what i'm hoping will happen is wherever let's say the points are sat on top of a blue unit or a red unit they will they will have a particular code that i can then organize later categorize the points by the the rgb code or something like that that i can then apply lithologies to in a much quicker quicker way um also what i'd hope with this is if i can work out how to do this this would actually end up um probably being a more defendable way of actually categorizing all of these points so effectively you know rather than me deciding what lithology they should be it's based on the actual the hard cold data that is um that i've been basing my interpretation on hopefully that makes sense okay and then pretty much the final step in this workflow and i know it's a quick quick overview through it but the final uh final step is to use the join attributes by location tool again available in the um in the uh the processing toolbox and we join the attributes of um of the point seed that we've just created with the blank polygon layer that we created from our from our framework what this allows you to do well it it allows you to uh join multiple attributes all in one go so that's really handy so if you've kind of uh if you've processed this you know if you've added multiple attributes to your point seed layer you're able to just process all of these straight into your polygons you can export this start to style it do whatever you need to do and start to build your map at that point and um and then you can sort of apply uh the more universal attributes that we would normally put onto our onto our files so that would be things like the date of digitization the um the digitizer's name things like that that apply to all of the all the polygons across the entire license area and you can do that simply by selecting the whole map and you know updating the fields so i guess the main benefit for this workflow uh that i found is um well firstly it's really helpful for large maps or intricate maps kind of like this where we've got multiple stacked units um and especially if it's going to need to be updated by i don't know ongoing work or something like that so kind of as an example for this project in particular we undertook a version one interpretation based on the satellite imagery and combining it with historical published literature on the area and other sort of information sources we were able to sort of project predict the main rock types that we were expecting to see based on that satellite imagery uh we then went to site and we undertook a couple of weeks of ground verification and general sort of like uh targeted mapping on areas of interest and as we were doing that we were able to update the the framework layer so the framework line the line work that we produced um and any kind of the point seeds that uh you know if we misclassified anything we were just able to update one or two of those points and then rerun that the entire map and it just it just completed completed the map and brought in all of our amendments instantly um the good that you know the quality of line work that needs to be done for this kind of map uh really sort of translates into the final map and i found this process to be much quicker than i guess what i'd class as a traditional digitization method which would be drawing multiple polygons and using snapping tools and things like that to try and get polygons to join to each other i think especially for maps of this kind of scale and complexity it really reduces the chance of getting overlapping polygons or you know gaps between polygons slithers and things like that which to be honest it's just it it starts to sort of show sort of poor uh digitization uh methodologies and verification um and to be honest once you've kind of got them in that system they're really difficult to then remove so i think using the line work to influence the rest of your map is is key to to producing a good good linework file and and a good map and the output finally there are a few a couple of little downsides to this as there always would be um it is it is slightly more time consuming maybe to produce your point seed layer especially if you're only doing a small area you know the larger the area the more i would recommend this this workflow to be honest you know small areas where you've only got to draw a few polygons you probably don't need to go through this whole process but if you're doing a you know a good size you know a good size uh interpretation then i'd really recommend this uh in addition to that uh i did find when i first started using this that there was a couple of glitches and errors probably more related to user error than the actual software itself but what i've been finding is the more that i practice with it and the more that qgis develops uh you know the slicker and the more streamlined this workflow is becoming so there might be a bit of a learning curve to get over first of all but uh once you've kind of adopted this format i don't think you'd go back to the other methods of digitization and um yeah so this is going to be my my last slide now so um some of you may have noticed i haven't mentioned cross sections yet now if um if you're unfamiliar with the geological processes if you if you're not a geologist a cross section is effectively uh it's a it's a slice through your map to try and represent the kind of what's going on under the surface based on your your surface observations so it's like a predictive it's almost trying to turn your map into a three-dimensional kind of uh interpretation uh qgis does have a cross-section builder and it's called q-prov uh i've just put a you know put an image of it up here so you can go and have a look at this in your own time um and again it's it's available through the plugin manager you just got to download it as a plugin and it's actually a really good little tool it allows you to define a section line on your map it will then draw a topographical profile based on the satellite imagery or your background imagery and then it will also display any intersecting structures and lithologies onto the surface profile so that's really helpful because you know um well effectively what you're looking for is to try and project those contacts um wherever they intersect the surface along your section line so it kind of automatically does it for you which is really helpful um i have struggled to sort of get it to do the entire process of building a cross section so what i've ended up doing is i get to the point where i've got a good topographical profile and i've got my contacts and a few structures and then i will extract that and put it into a um like a digi a digital drawing software package for example um adobe illustrator is good for this or if you're looking for an open source uh uh an open source option inkscape is also really helpful so that's how i've drawn this cross section here it's quite simplistic but um but yeah you know i guess the more time you put into it and the more detail in your maps and things you're able to start to really um you know you you could really produce some beautiful cross sections using this methodology i think um it'd be nice to see that you know i think coupon is a little bit clunky so it'd be nice to see if that did actually develop some more functionality but to be honest it kind of does what i want it to do and this part of the workflow is is something that i'm still actively working on so um yeah if anybody's got any insights and has got other ways of drawing cross-sections in qgis i'd be really interested to hear um but i think again with a bit of um you know a bit of a learning curve and stuff you could you could start to produce some good digital uh cross sections through this process one thing i did forget to mention at this point i guess is um is that actually once i've drawn my cross section i would then bring it back into qjis as an image when i'm starting to build my final map layout output and drop that in underneath the map and be able to display it that way so you can still include it in your final map layouts and things like that um so i know it's been a pretty high level quick run through through this workflow but i hope it sort of stimulated a few ideas for you uh for your own processes and um yeah if anybody's looking for any more information or wants to discuss this further i will um well i i was going to try and drop my um my linkedin page in um in the chat but i'm not sure if i've got access to that at the moment so um yeah my name's chris lambert i work for slk it should be fairly easy to find on linkedin please feel free to get in contact and be happy to discuss things further with you uh thanks very thank you very much for listening uh happy to take some questions thanks thanks chris um once i put up the video at the um end of the um stream i can absolutely put your contact detail for linkedin in the comment of that video so um just send me that and i'll be able to put that in for you there are a couple of questions and comments and advice so i'm going to start off with something quite simple that's more an opinion question and then we can build up to the other things so um dnadventure asks is qigs qgis as powerful as the esri product for geological mapping um okay so i would it depends on what you're trying to do of course you know uh i i've used arc gis i've used mapinfo uh as i said i've been using gis software for about 10 years um i personally i get on far better for 90 of my work in qgis uh it reads all the shapefile formats and everything else and i've even started to find some plugins recently that will start to read esri style you know style files and things like that so you can almost translate most of your data across pretty quick and easy um there are still some features in in the esri products that um that are probably more powerful that you know it's horses for courses really it really depends on what you're trying to do like i say 90 of my workflow now probably 95 of it is in qgis and i just dip into the arc programs if there's something really specific that i need to do all right brilliant um and then wendy adrian asks uh how do you get the liniment and structural above the sentinel image do you have to use another data or tools the structure the structures um so i i manually digitize them so that is part of the interpretative work there isn't like well there might be a way of producing these lines automatically but i think a lot of geologists would normally agree that they'd like to do their own interpretation um so i just utilized that that underlying data and you know it doesn't necessarily have to be uh satellite imagery either this work this this workflow would work perfectly fine with any geophysics or anything like that you know so um yeah you can kind of um it is a manual process it's still the same kind of form of digitization but i think the main time saver is where you're digitizing a single line and you're not duplicating efforts by drawing polygon after polygon after polygon you know and you make simple tweaks to the workflow uh to the to the uh the framework file and then you just re-run the process so yeah i think it saved a reasonable amount of time in my workflows as i've been as i've been doing this yeah hopefully that's answered it i hope so and then um [Music] the sticker i i hope i'm saying that right um says semi-automated classification plugins or classification for this workflow or only for downloaded data question mark okay uh i think i understand right so slk we have our own internal methodologies that we use um so i haven't dabbled with the um the actual plugins in qgis that much um however i have sort of started to research into it just to see if it could help us with some of our internal workflows um the srtm downloader uh is a is an option to download satellite imagery um you can either do it directly through the satellite imagery providers um i've you know if you if you google for you know sentinel 2 or asta data you can find them quite quickly i forget what they're called right now but um the srtm downloader will also allow you to connect automatically to those services i think the time scale in downloading that data is it's slower in qgis than going directly to the provider but if you know it also helps you do some basic processing as well and like i say there's a there's a load of youtube videos and things like that on this subject that um you know the people who are posting those kind of videos are way more qualified about talking about this stuff than i am to be honest so uh yeah the the semi-automatic plug-in um that again that takes a bit of sort of getting used to tons of information online to have a look at but um that one i think allows you to start to sort of classify uh different type you know um different spectral responses and to you know sort of effectively give those criteria so for vegetation i think it's really useful for identifying different types of vegetation you can give it like sample subsets and it'll identify all of the the rest of the data across your your license area using those kind of those training points effectively so um yeah again probably best to just go direct online and have a look at the people who are sort of doing this day out you know i normally just use the data for my geological interpretations rather than doing too much processing myself so yeah all right and then there two sort of general comments um i think about at the point where you were asking the best way to sort of sample the values on the underlying image nazar's um says raster analysis and then sample raster values um i have put that in the comments of the jitsi if you open those yeah if you open up the comments i've put in another comment um and you can see by champlain i'm assuming and then he says why sample the points in risk hitting noise data can't you get mean values of the raster data below um and then the respective polygons or within the respect but well i guess that's kind of why i wanted to give this presentation because you know if that's possible then that's certainly something that would be really useful i think so uh yeah i mean if so that might even remove the entire need to do the or um uh your point seeding you know if you can just generate your polygons and then classify them based on on the raster responses underneath those polygons that would be awesome so uh yeah uh that yeah like i say it's a work in progress if that's possible that'd be awesome all righty i have a couple more questions for you um wendy again um she asks is it better to have landsat or sentinel um for structural interpretation okay uh we typically would use uh i think we we normally use sentinel 2. um again i'd have to talk to i guess we're going to get more into the realms of remote sensing here rather than sort of map digitization so i guess um i'd have to have a look into like the actual uses of particularly um you know satellite missions over others i know that you know based on the hour in in team guys they normally recommend the sentinel 2 for a lot of this stuff but i know that we have used landsat we normally use aster to do some mineral mapping um and also if the project's got worldview 3 the quality of view two or three i can't remember which one is but um the quality of that is is above the sort of freely available data but you have to you have to pay for that information so some projects we have really like like amazing satellite imagery and other times we have to make make use of what's available um you know freely available online so another data question um what dsm data is more accurate srtm or alos alos um thank you that's the question do you happen to know i don't think i could um [Music] i i don't think i'd be able to comment on that with any confidence that is all right um and then just a lonely job a question that i i wanted to ask this might just be totally you know left field but um for your um could it be possible to use specific like the specific color bands that are reflected by the rock types to classify them so in that raster you have those those color bandwidths i know we do it with vegetation mapping a lot could you could you use that to maybe create like a polygon layer are you able to do that yeah so we would um there has been a kind of a look into how to do this and um i think the kind of the general sort of feeling and correct me if i'm wrong if someone else in the audience knows better than i do but um with vegetation it's fairly uniform whereas with rock masses it's um it can be very variable especially like even even from one side of your license to the other you know it might be the same geological package but it might have might have a slightly different spectral response we also when we're looking at rock type there's also a lot of other factors i haven't discussed in this in terms of um you know that we have uh alteration so the closer to those granites you get the more altered those rocks will become um where next to faults there's alteration uh there's a load of other kind of factors which can really sort of jumble up the signatures so um it really does have to be a kind of manual process but we're sort of trying to automate certain portions of it um i think you know in terms of um yeah in terms of sort of generating polygons from that sort of stuff it's it's something that we're kind of working on but we just haven't got anything that's really been able to sort of be nailed down just yet so yeah all right um tim do you have anything more to say uh no maybe just a suggestion to go and have a look at the process the modeling tool that you can create models with the workflows since you're using lots of processing tools on their own you can actually string them together and create one continuous workflow and otherwise very nice to see what you're doing and entertaining thank you for the presentation that's great thanks tim alrighty um i think as far as that goes um i will put up the video with the questions and that sort of thing and link your linkedin to that so if anyone has any further questions or advice for you they'd be able to do that yeah that'd be great there's just sort of every now and then i look at the comments and they keep coming keep coming um i just uh quite a lot they're complimenting you on your workflow and um that this is this is really cool and i think it's really amazing as well you know i think one of them the coolest things that has been shown in today's open day is how diverse cue just like how diversely it can be used for so many different applications from chris's bob ross painting this morning to making hardcore geological maps it's just a really powerful tool yeah and i think it is i think it's just great that this there's this kind of thing and this community that's you know that people are willing to help each other out and show their workflows and and you know i think everybody can can you know swap ideas and and develop themselves is a brilliant tool for bringing people together i think so yeah one just one last question um is there any method to evaluate the certainty of your map no um okay so i i have done a couple of projects which have kind of started to try and incorporate some kind of machine learning kind of aspects to them uh and you know uh sort of trying to put a number and a value on certainty or should we say uncertainty would probably be a better way to um uh again it's one of those things you know i'm i'm not a i'm not a python user you know i'm kind of a self-taught gis user i've had a couple of basic training courses and stuff but pretty much everything i learned is off youtube and um web forums and things like that so um i have seen a couple of people doing this but personally um i haven't i haven't delved that deep into it yet it would be a really nice thing to do actually though um and again yeah i don't know if if maybe that if somebody knows something about that that'd be a great talk for another one of these kind of sessions and i'd be really interested to have a look into that yeah alrighty um there are last couple few comments just giving you some advice and i will copy those into you into you um although they do say that in the processing toolbox under raster layer zonal statistics um is the way to go to sample raster values um for your polygons oh that's brilliant thank you well i know what i'm doing this weekend i think we all know oh brilliant all right um i'm gonna round us off there i think we've had quite a fantastic um day um of all of our different sessions and i do want to thank everyone who's been watching until the end of our um open day and we're very pleased to have been involved and to have organized it and just to see how many people use qgis in such a wide and diverse range of um subjects and um thank you all for watching and we're gonna head off thank you amy and the whole organizing team uh it's been fantastic to watch you all get on and put this day together and looking forward to doing the same again next month thanks tim yeah good job great thank you everybody

Channel: QGIS

Views: 11,761

Keywords:

Id: FpDrm985Tn8

Length: 44min 20sec (2660 seconds)

Published: Sat Feb 27 2021