Zonal Statistics and Zonal Statistics as a Table

this video will review zonal statistics and zonal statistics as a table zonal statistics found conveniently under the zonal toolbar can really be thought of as the spatial joins of the raster world where they allow you to take some zone or some grouping right so maybe the grouping is just all farms maybe the grouping is every single farm individually maybe the grouping is a vector layer here like municipalities in Lancaster a guard 'less of what it is those zones those boundaries those discrete contiguous areas are calculating some statistic that's interesting just give that a sec sorry about that right any of them is going to be calculating a statistic based on some raster grid right so I have a grid cell or grid layer of elevation here so maybe I'm gonna have every single farm right tell me what's the average elevation for that farm or what's the rate of change the you know the range of elevation or what's the max elevation same thing maybe I'll have each municipality tell me right what's the range of the elevation of what's the minimum or maximum elevation alright so the first thing is almost statistics as a table so opens oh no statistics as a table what comes first is always the zone right notice raster a feature right so I could input a region group farms right every single individual farm or what I want to do you need all right muni municipality right i take a look at where the values are these are the values that i want so the zone field is municipality essentially zone field is what's the field right or the value if it's raster that's causing these to be considered discrete right in this instance i'm saying your discrete based on what your name is if your Adamstown borough you should be diff than accurate burrow so that's what that does value raster is what's the raster I'm using to calculate statistics we want to do some statistics for elevation almost say municipal zonal there's almost at statistics type was almost exceeds the table I can choose from many right the average the maximum the minimum the range the standard deviation the sum right one thing that it will notice here rather I'll show an example of this in a second right but let's just calculate all statistics here it okay and I get a nice table for every single municipality all right those are the zones and they tell me order that but the maximum elevation in each is the range the mean the standard deviation and because this is a table I can take it and I can join it right back alright so join it to that it was municipalities municipal and now every single region can tell me all of this information area that's pretty neat and so maybe one thing I want to show is the range right show me which ones have the greatest rate of change so I'll go to quantities pick range right set a quantile and now I can actually look and lo and behold if I jump back and forth those do tend to be the ones that are in cata hilly areas where they rapidly change right from kind of Hills to not where these tend to be areas that don't see a lot of great change I'm gonna remove the join all right and I'm gonna bring in land cough in something that we've been using in class occasionally an Atlanta cough in if you remember really all it is is its layer saying a value one through five for whether you urban agricultural forest wood or other right these are meaningless values in the fact that 2 is not greater than 1 and 3 is not greater than 2 they're used to almost categorically classify whether something's urban agricultural forest when I used elevation the numbers actually mattered right like ten is two times as much as five and twenty is two times as much as ten so whenever you want to calculate some statistic here I do zonal statistics as a table ringing land cough it right because the numbers are integers now and they're not float like they were here it opens up new ones we haven't seen yet primarily majority median minority and variety eight majority which value happens most right so if I clicked majority it would tell me for each municipality which land-use type is occurring the most right are there the most cells with the value one two three four five same thing with minority what's happening least median you know what a median is it looks at them all and looks for the kind of median value that wouldn't necessarily be helpful to us here because one two three four five they're not kind of numeric values rather they're not linear values and finally variety right variety it would give me an answer between one and five and what that would mean is there's one right type of land use in it or two types of Flanders or three or four or five we'll do majority because it's the easiest to understand right then we open it up very cool you can see a lot of urban a lot of farmland some that happen to be for us nobody has wood or other as their major use I can join this back to the municipality where I do that let's just change these values ain't bourbon yeah little gray agriculture will be a light yellow forest and shrug be a dark green withering wetland be blue other be all right so I'm gonna join what I just created munis majority sorry municipality was the zone for each now each municipality is telling me by the land-use type that happens most right one your member means urban to means agricultural three beats forest all right so let me symbolize them in such a way that I can see for each does it have more her balloon does it happen on farmland or does it have more forest land and so I'll take municipality symbology and go to categories right and I know that means urban agricultural forest let's try to give them colors that are similar back to what we've seen before the grays the yellows the greens and look at that I can jump back and forth and I can see I'm pretty accurate right those municipalities that tend to overlap the largely urban areas tend to have most of their land earth a lot of areas tend to be farmed but there are several areas here here and here which tend to have a majority of the land cover in their area as forests alright so we've seen its use as a table and joining it back to a vector or raster but you can also use it as a raster right so what I'm gonna take here as my zones is the region that I did a couple steps ago totally arbitrary value all that it means is I am a discrete contiguous farm block cool just an object ID sort of so I'm gonna say for those region group farms right the value is what differentiates them and makes them discrete I want to calculate some statistic about elevation we'll just do rain because it's easiest farms that rage it okay and what returns is a raster and look at that it's values that essentially mean the range of elevations and so I could symbolize them in such a way that they're going from low to high here and all those areas that have a high right range of elevation change if I look at where they were look at that they're lying right on elevation where there does appear to be whereas these ones appear to be in areas that doesn't have a huge elevation change alright so the difference between the two is when I use zonal statistics my output is a raster and the values are whatever statistic you calculated about this grid layer right so I had farms their kind of value that determined are they a discrete farm that's what treated them as zones they calculated some statistic about raster elevation sorry they calculated some statistic about elevation which was the range in elevation and they returned with values that meant range in elevation right so I could go to that the range in elevation here is 137 205 109 234 94 181 278 and so on

Channel: Shea O'Neill

Views: 32,242

Rating: 4.8684211 out of 5

Keywords:

Id: S3Bj_SWMvBo

Length: 10min 25sec (625 seconds)

Published: Thu Oct 16 2014