GeoServer for Spatio-temporal Data Handling With Examples For MetOc And Remote Sensing


Formal Metadata

GeoServer for Spatio-temporal Data Handling With Examples For MetOc And Remote Sensing
Title of Series
Aime, Andrea
CC Attribution - NonCommercial - ShareAlike 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this license.
FOSS4G, Open Source Geospatial Foundation (OSGeo)
Release Date
Production Year
Production Place
Seoul, South Korea

Content Metadata

Subject Area
This presentation will provide detailed information on how to ingest and configure SpatioTemporal in GeoServer to be served using OGC services, with examples from WMS and WCS services. Topics covered are as follows: * Discussion over existing data formats and how to preprocess them for best serving with GeoServer * Configuring SpatioTemporal raster and vector data in GeoServer * Serving SpatioTemporal raster and vector data with OGC Services Tips and techniques to optimize performance and allow maximum exploitation of the available data The attendees will be provided with the basic knowledge needed to preprocess and ingest the most common spatiotemporal data from the MetOc and Remote Sensing field for serving via GeoServer.
Server (computing) Process (computing) Open source Software developer Projective plane Expert system Number Array data structure Computer animation Linker (computing) Core dump Computer network Reduction of order
Area State observer Medical imaging Computer animation Presentation of a group Scientific modelling Multiplication sign Set (mathematics)
Computer file Presentation of a group INTEGRAL Multiplication sign Scientific modelling 1 (number) Shape (magazine) Raw image format Mereology Number Medical imaging Bridging (networking) Database Core dump Cuboid Data structure Physical system Form (programming) Social class File format Moment (mathematics) Bit Limit (category theory) Subject indexing Computer animation Personal digital assistant Hausdorff dimension Spacetime
Point (geometry) User interface Computer animation Personal digital assistant Weight
Medical imaging Subject indexing Computer animation File format Cube Hausdorff dimension Element (mathematics) Weight
Computer animation Information Concentric Multiplication sign
Computer file Multiplication sign Projective plane Content (media) Expert system Coordinate system Inference Cognition Web service Message passing Computer animation Database Configuration space Physical system
Revision control Message passing Product (category theory) Computer animation Cube output Website Function (mathematics) Weight Raw image format Form (programming)
Subject indexing Computer animation Computer file Demo (music) Information Database ACID Mereology
User interface Multiplication Service (economics) Demo (music) Computer file Multiplication sign Scientific modelling Content (media) Representational state transfer Weight Software maintenance Trigonometric functions Subject indexing Computer animation Cube Extension (kinesiology) Window Library (computing)
Gaussian elimination Computer animation Query language Cube Multiplication sign Hausdorff dimension Range (statistics) Cycle (graph theory) Parameter (computer programming) Client (computing) Quicksort
Pixel File format Distribution (mathematics) Polygon Physicalism Set (mathematics) Line (geometry) Sequence Inference Number Array data structure Computer animation Visualization (computer graphics)
Computer virus State observer Musical ensemble Computer file Connectivity (graph theory) Direction (geometry) Combinational logic Complete information Point cloud Parameter (computer programming) Event horizon Order of magnitude Array data structure Natural number Flag Arrow of time Extension (kinesiology) Social class Physical system Area Product (category theory) Projective plane Set (mathematics) Computer animation Integrated development environment Visualization (computer graphics) Personal digital assistant Order (biology) Module (mathematics) Website Quicksort Routing
User interface Default (computer science) Computer animation Network topology Cellular automaton Flag Mass Active contour model
Point (geometry) Server (computing) Computer animation Archaeological field survey
Server (computing) Information Multiplication sign Projective plane Exponentiation Sheaf (mathematics) Parameter (computer programming) Function (mathematics) Mass Field (computer science) Subset Computer animation Operator (mathematics) Code Cuboid Quicksort Data structure Escape character Extension (kinesiology) Communications protocol
State observer Thread (computing) Java applet Multiplication sign Modal logic Combinational logic Parameter (computer programming) Weight Subset Web service Extension (kinesiology) Descriptive statistics Physical system Covering space Email Product (category theory) Structural load Gradient Flow separation Photographic mosaic Data model Graph coloring Hausdorff dimension Duality (mathematics) Right angle Quicksort Uniform space Data type Resultant Point (geometry) Server (computing) Presentation of a group Computer file Mass Raw image format Goodness of fit Latent heat Energy level Friction Form (programming) Addition Information Content (media) Volume (thermodynamics) Local Group Subject indexing Computer animation Personal digital assistant Cube Matching (graph theory) Library (computing)
Computer animation
OK thank certifying Henningsen then I'm into they will speak about n-dimensional the handling induce earning particularly that M an accent on the rest of the data as opposed to a vector it so I would produce solutions your solutions the
company based in the link founded the late 2006 we are imagined processing experts and uh contributors had a contributes reduce server due tools GeoNode as you network and so on so we in battle involving a number of open source projects and in particular I am 1 of the core developers of GeoServer tools so that has
an area for this presentation his that uh while you have a lot of n-dimensional data that's coming from remote sensing from a forecast models in the from seat observations and you want to displayed to make it available and possibly underlies it would use so as a
set of going to talk mostly about the rest of the time so way to publish and they mention of it and users is to use the image most like the
image was like is as our is 0 aggregation of image files of I believe distributed in space and time in all the they mention if you want to we have an index of indexing all these files and economics contain them it's relocation time innovation and other they mention of Dutch that particular image and putting them all together we get an n-dimensional so we have a number of assumptions on the Gunnison there bits that make up there some MOS like 1 is that they they are all in the same coordinate the reference system so you cannot MOS I together the bridges 84 you and the teacher north at the start of the moment we are planning to remove this limitation uh we used to have their requirement of having all the images in the same column model so you could not mix together gray in the RGB and collected but introduce eventuated these limitations be removed of runs can overlap as we please we don't care so they they can be nicely aligned or partially partially overlapping there's no problem and they can be in different format if you need to also normally and was like is made of fights all in the same form as 7 was like index is the core of the structure it the structure that we use to quickly locate today defies the granules that we needed to answer ascertain request about certain bounding box at the same time as the innovation the index can be implemented in various ways we can have a dating that's a stored in databases special that abuses such as Paul Oracle at you but can even be just a shape for in case you have to integrate with the legacy system you can implement an interface and integrate with whatever about your cost of the the major it's if you to the mentions the dimensions are well normally in and mission was like they are tiny relation but then you have a number of excellent limitations such as their fair customer and the last update at the time of the file and so on so we can have look all although there's a dimensions and all the extra ones without without problems all of them are stored as reducing the index in
that in case you have lots of lots of files in the yard convention where they're dimension value is part of the findings we can have them was like a court your so that class is the dimension out of the file name and puts it into the next week so
once you have the two-dimensional ready you can go into the user interface and enable them in that case they will show up in the capabilities document of problem as the Lucius uh and other point declines will be able to use them and factually that Italy once now n-dimensional is
normally connected with net CDF usage of net CDF is up but only
get the most common multi-dimensional data format that you can find so you know what normal duties normally contains only 1 image net CDF user with mission of format the thought to provide the uh cubes of data n-dimensional cubes of data so you have what 1 master which is not flat but it's actually the 4 5 dimensions and uh we talked to your and you close to read it and to expose it as a coverage or as more than 1 college if I
need to uh net CDF is not exactly the fastest format to to read the because it's and fermented to perform that exchange but we created the some side indexes that we used to speed up access to a particular element of their n-dimensional cube so that we can quickly access and look at the data that we want without having to really build the which makes it suitable for on-line publishing even if the former was not exactly the size of the the a and that CBS's strange also in
another way I 1 find contain multiple information I can have the concentration of ozone and nitrogen and other gasses in the same 5 so have multiple covered using 1 5 is something that you saw it did not know of parental but we extended it so that it can in
uh nested if is also pretty pretty complicated inside there are conventions they are called the climate and forecast conventions but 9 times of the time when you get an entity and find that claims to be CF compliant it's not and on slightly beating the time that it was not named right away the elevation uh is not from 0 but it does a Bayes value and well it's a mess so we have our
and unexamined file configuration file that you can edit and I mean if you're lucky you'll if you can get there the 1 that's all generated but service so you don't have to right it but if you want you can use it to how would you server understands the contents of the message if I may be throwing away 1 coverage that you want to want to sheet and to show outside or that it's that that the times are actually be used on the on inference times the the so the
scientists playing with Matt CFR not like very much the EPG database and that the 5 thousand coordinator for a system that are in there they want to use their own so there thousands of extra cognitive systems that the scientists use and we need to to find a way to publish them would you server so you can create your own little database of expert according to of a system that you can feed into GeoServer and faster than to model so that they get automatically recognize published without having to pull Harris every time with a different projection
the so uh essentially annotations about and that's again well uh in earlier version of you so that you could only only support the message if fights in the rigidity form but that's just said no we support whatever so induce eventuate we can have a mermaid as may productions as you want we want to the find to be CF compliance roughly and if you want to I'm at CDF output from Juicer it you have to use that Lucius so which we don't do fall on the BMS for example as an output input is fine now we can
put them together written at CDF and handling with MOS IQs because in many mythological institutions you have won a CF well worth of of fire what of data for 1 day and then you have another and then you have another so you have homozygous net CDF site you are more was I King of some n-dimensional cubes in practice we have
extended demos like so that it can handle this kind of situation and then we have we basically integrated the index of their that's Jeff when the index of them was like so that we can quickly locate which snapped CDF file in which part of the acid if I want to read the when answering questions to do that we also added to
extend the dead being Teixeira all fall off from of the malls I could a against an XML file to declare all that they mention also the G is the the schema that you you are going to use in the database to store the information about the vertical college and so on so I he grew a bit complicated but then again it's is complicated so we could not avoided I once we
did that uh then was like was able to publish multiple ecologies multiple n-dimensional ecology is out of it so we extended also that user a user interfaces so that you can see the multiple colleges coming out of them was like of net CDF stand aside which 1 to publish now in all the
scenarios you normally try to maintain a moving window of data so you have an n-dimensional cube for 1 day and then another and then another and you probably want to keep the last 2 3 months of data war uh online so every time you add a new day you want to throw away another day so basically that we need to have a way to automated that maintenance of these malls like over time so that we don't have people going on the user interface or in other ways to update to the contents of the index so we created the REST API dedicated to them was like in which you can query the contents of demos Iike locate the size that you want to remove you can move them you can upload a new files and they would be harvested in index said the into the most like so this gives us all the price of seamless support formerly window those are the cosine of going to get into details I talked only about net CDF but in fact on we also supported with at the same genome-wide library that can be helps us reading that also helps us reader upgrade so of whatever we can do with Messier Pfizer so indexing them and putting them in animals I can publishing them we device services we can also do we grieve the only catch is the grid is not as well tested as net CDF in fact if you if you check support is on extensions of publicly supported would fight is a community model so it's available from nightly builds but it's not officially supported yet because we didn't Triton now that we know that it doesn't work it's just that you can try to now all this
data available how do we surveyed out to the work I I have have of course
in that less so I have them all cycle for all of all their and their Michelin cubes spend as you've seen before indicate capabilities we advertise the range of elevation that range of times that range of custom dimensions that we exposed sort client in me I get my request with the some extra parameters this is called Dublin nest he promised temporal but it's not just temporal it's in that mention the fact and here I'm asking for a particular time a particularly nation into custom dimensions that dimension are kind of strange they have this deem underscore pyretics before then it's by spec to avoid confusion with other parameters and so here I am all query Essex they mention deficits chool sponsor eliminations plus the for of specified here of course uh seen in because
there are that he's nice we assign a column up to the age of physical bodies and when you see the distribution is nice but
sometimes you really want to the sequence to lines on the polygons all values between 0 1 and other or if you we the you want to see when bodies and so on so we got this rich data set and we would like to expose it in a number of visualizations you several we have the concept of our and the inference formation under house formation is a way to transform the because as rendering so we start with pixels and we might end up we conclude line expected on the fly and then we think them as vectors so we can do visualizations like this
thus the class contour lines competent on the flight fast because they are comforted by solution of which I'm looking to the so sun not to nature solution and only inside the the area the looking and uh this is a way environment and that sort of c currently involved in those cases we have their cottages which you there's normally they are either well at the you in the component of the vector of the wind so the projections of the vector and the 2 axes and then you have to do so my thing your site to generate a magnitude and direction of the arrow and I think the article you can see we have a we virus which are multiply and then we have a little tricks in the if you have complicated
products we also have a module called the Dublin messy all uh this is out of the system the uh at observation extension formidable messed and it helps you to publish a complicated that set of is a free you have there the route which uses of housing at and then have you know we you can have outlines which are their outlines of the files composing your your product which have a vector and then you might have their all bands of the product and you can ask for a combination of events so you can do for score visualization you can have it you parameters which are are processing on top of their the base data that you can display and you can have the flags which are the presence of clouds presence of his no aware this season's on some area in which that it could not be collected the song from each subvector and roster the semester are binary Malaya certain all orders are complete information we have our complete
user interface to build this tree because it's rather complicated to put it together so that it's actually compliant with this fact you have to set up like then default styles for each cell for the flags of the contours and so on so we have a user interface guiding you through setting up this then the idea of the day when you are
published this you can just go to the Preview and look at your data and I have compliance flies that will depend on the search for what it I spoke about valid the mass is
not about rote surveys so what we do when we want to download all data we use the
yes in particular the CS 2 point all which is implemented in just server is particularly well suited to download n-dimensional later that was yes to all
is sort of very pluggable tropical like you can make a that was yes 2 . 0 compliant server supporting just bounding boxes no they mentions noted projection on nothing are and then you can show people what you can as extensions while just over basically picked all the sections so you got everything you need you cannot escape you Canada project that and then you can control how the encoding is done for example in duties you can decide whether or not to reject conference the output or how they unifying structure would be by just adding some extra parameters to the coast and then
just as described college operation that we use just server to provide information about all the the resistibly all by itself would give you information about time and elevation but in just a we support custom they mention we support them in Dublin mass we wanted to support and I mean the busiest to so that you can a subset of the exponential as well if you want to well there would designed to the BCS poblano protocol allowed the coffee section of their described to be field and we are finished with extra negative information that you can use this is described in the nation the way of in and and so on and at
that point you can be sure this 1 is this long very long duality is on the cover to rest the recent some subset over several dimensions so longitude latitude in addition telling in there I'm getting out out an n-dimensional cube which I can only representing a so I can ask the user to encode the opportunity in yet out a Q of information as opposed to outflank last and all the explanation that we support in our describe coverage are supported by as well finally we also
support I mean that the of sufficient yeah observation extension is again an our extension on top of the Bayes specification that allows to describe a complex products which are made of many uh files so in our case the image mosaic image mosaic is made of many times Mg we publish it as I have observation that was in the result of described the all the friction that we can use to quickly browse the contents of the moles like uh this is very nice because when you have several dimensions is difficult to figure out which combination of the values of their dimensions of the match is 1 of the files that you have in store and in this way you can actually see them and then locate they're the right combination of parameters loads of and that's what the yes what's more time for questions what questions the the yes this it was the group of that the common data model this year maybe I the low level and I didn't work on it directly I just know that the user you cut libraries to tools to pass the necessity of in grade but to no idea with your color what group supported me if you're if it reduces to the M DPI and at any uniform at that it's a Java supports will be supported good evening register raw thread servers so my descriptions of the unfortunately I I don't I don't get it down to that level I I worked on this service spots and then you know manually Michaeli which is 1 of the people that made a presentation at work in the lower levels of sort of apparently don't know that because this is worker in the US we have threads in Europe there used to the system so it's kind of merging of the I know I know of masses lots of people are interested because of that great thank you whether the performance the kind forms would the dancers flocks facts the 3 depends only on the type of volume you were talking about so let's see if can grow begins to ways that significant grow bigger because it has a army yellow combinations over the possible emission values so you have a muon sliced but because they believe that the small so it's I don't know mn plagiarism by 1 thousand yet and I have my own type of and and therefore that is our external index helps a lot because we can basically jump directly to the offset the doctor has the data without the reading the header of the city and so on so that cases fast and then you can have very few dimensions but a very large greed and in that case that there's not much we can do because the main problem is that it's a does not support over use How happily most of the net CDF mythological data which is which does not have uh haida solution it normally has a ton of dimensions to solve for the sweet spot we are good therefore the case in which you actually have has 1 million by 1 million peak system I would try to use something else thank you and was in can I ask is is mentioned
here the 1st speaker 1 of 20
minutes and the


  567 ms - page object


AV-Portal 3.11.0 (be3ed8ed057d0e90118571ff94e9ca84ad5a2265)