SkinnyWMS Meteorological Web Map Service
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 351 | |
| Author | ||
| License | CC Attribution 3.0 Unported: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/69196 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language | ||
| Production Year | 2022 |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
| |
| Keywords |
00:00
Self-organizationWeb serviceState of matterInternet service providerState of matterWeb serviceOcean currentSoftwareWordSelf-organizationReal-time operating systemRange (statistics)Computer animation
01:18
SoftwareWeb serviceOpen sourceComputer animation
01:47
Visualization (computer graphics)Workstation <Musikinstrument>Different (Kate Ryan album)PredictabilityProfil (magazine)Time seriesParameter (computer programming)Multiplication signCoordinate systemExecution unitMetrologieCASE <Informatik>Level (video gaming)Arrow of timeExpert systemDimensional analysisVisualization (computer graphics)Computer simulationState observerVertex (graph theory)Point (geometry)Form (programming)PlotterWordType theoryRaster graphicsNumeral (linguistics)Presentation of a groupComputer animation
05:22
Range (statistics)Table (information)MetadataStandard deviationFile formatCodeAdditionSet (mathematics)SoftwareWeightParameter (computer programming)Multiplication signPoint (geometry)Term (mathematics)Level (video gaming)MetrologieVariety (linguistics)File viewerExecution unitDevice driverVisualization (computer graphics)Representation (politics)Interactive televisionOpen sourceComputer animation
07:17
Stack (abstract data type)Operations support systemCompilation albumCartesian coordinate systemProjective planeTesselationInterface (computing)WeightComputer fileWeb serviceOpen sourceCodeSoftwareAdditionWeb 2.0Level (video gaming)Computer animation
08:17
Maß <Mathematik>Range (statistics)Internet service providerWeightFile viewerWeb browserDynamical systemVisualization (computer graphics)Execution unitLatent heatParameter (computer programming)Dimensional analysisMultiplication signData conversionField (computer science)Online helpMetadataDifferent (Kate Ryan album)MetrologieWorkstation <Musikinstrument>CuboidServer (computing)Point (geometry)Degree (graph theory)Computer animation
11:04
Personal digital assistantData storage deviceProgrammable read-only memoryComputer animation
11:31
State of matterRoute of administrationOpen setWeb serviceClient (computing)Time seriesVisualization (computer graphics)BitDifferent (Kate Ryan album)SoftwarePoint (geometry)CodeFile viewerLevel (video gaming)Profil (magazine)Traffic reportingPlotterSinc functionExecution unitRight angleVertex (graph theory)Software bugMereologyWorkstation <Musikinstrument>Parameter (computer programming)Open sourceMultiplication signInformationStandard deviationWeb 2.01 (number)WordFile formatData conversionTerm (mathematics)UsabilityDegree (graph theory)Mobile WebServer (computing)MetrologieWeightComputer animation
15:15
Web serviceCartesian coordinate systemDenial-of-service attackContext awarenessData storage deviceSinc functionOpen setComputer animation
16:00
Open setComputer animation
Transcript: English(auto-generated)
00:00
Thank you very much everyone for coming. Thank you for the introduction. My name is Edouard Rozert. I work for ECMWF, the European Centre for Medium Range Weather Forecasts and today I have the pleasure to present to you one of our software which is called SkinnyWMS.
00:22
So first a few words on who we are. ECMWF is an intergovernmental organization. Currently we are comprised of 23 member states and 12 co-operating states. We provide not only real-time operational weather prediction service to our members
00:43
but also we are a research institution so we continuously improve the numerical weather prediction but also provide climate reanalysis data. And we currently operate two EU Copernicus services.
01:00
One of them is the Copernicus Climate Change Service and the Copernicus Atmospheric Monitoring Service. And on top of that we have a long history and experience in providing software to process meteorological data and also visualize it. And one piece of software that I'm presenting today is called SkinnyWMS.
01:26
So as you can already see, I hope my animation works. Yeah, seems to work. So SkinnyWMS is a meteorological web service. So it is an open source tool that allows you to visualize meteorological data interactively.
01:47
But before I go deeper into what kind of features SkinnyWMS has, first a few words on the challenges in meteorological data visualization. So the two most common types of meteorological data are rasterized data and point data.
02:11
Rasterized data, usually you get it in the form of numerical weather forecast data. So this is quite common. Point data usually is station data.
02:24
So it's individual observations of stations all around the globe. But also it can be simulated station data as a kind of prediction of future values.
02:40
This data usually is very diverse. So we have a plethora of meteorological parameters with different units. It's not necessarily the case that the same meteorological parameter has the same unit or is provided with the same units depending on who is producing the data.
03:03
Also sometimes some data is summed up over time. So you need to understand how actually these values came to be. We have categorical values which a good example is the present weather parameter which basically for example can show you with a numerical weather
03:26
thunderstorms are present and things like that. Usually when we talk about meteorological data it is time dependent data that can have multiple spatial dimensions and depending on who is producing the data
03:42
we are dealing with a lot of spatial coordinate systems that the data is provided with. Also some of the challenges are numerical weather prediction data as I said is usually coverage data so you get basically a raster all over the world and this might pose challenges
04:03
for example when you want to visualize wind so you have wind arrows that if you present them on a map they should ideally not overlap so data thinning is also an issue. And of course if you have multi-layered data transparency and things like that become an issue
04:24
so this is already quite a lot but traditionally meteorological data was used of course for meteorological experts or for forecasters so over the time different specialized let's say expert visualizations were created for example a good example are
04:45
wind barbs so these are not simply wind arrows but also show the wind speed encoded in in a certain manner and this might be really challenging to actually visualize. And yeah as I said time series data means that we we can actually produce time series plots
05:07
vertical profiles if some of you are pilots then they might be familiar with vertical profiles of wind where you see different wind speeds and wind direction in different altitudes.
05:23
So as if this wasn't enough two very common formats to provide meteorological data are GRIP and NetCDF which also can be quite challenging to decode. GRIP is a standard managed by the
05:41
WMO so it's a WMO standard which is commonly used for numerical weather prediction data. It gives you a very compact binary representation but depending on who is producing the data it can be challenging to decode because it requires additional tables so these are
06:01
some of the challenges with NetCDF HDF5 you have a wide variety of tooling available so this is already quite nice but at the same time you're less standardized in terms of parameter names and units that that are used in these data sets some of it is mitigated if NetCDF
06:24
data follows the CF metadata conventions but all in all can also be quite challenging to actually understand the data. Luckily ECMWF provides some nice software tools for example EC codes or CFGRIP driver for x-ray which makes it already much easier to ingest this kind of data.
06:48
So now that you've heard some of the challenges with meteorological data let's get back to skinnyWMS. As I already said it provides an interactive visualization of the meteorological data
07:02
and it is already packaged to include a nice little map viewer so you can browse the data the only thing you have to do is point skinnyWMS to the actual data and in terms of technology skinnyWMS heavily relies on our open source software that I already
07:27
mentioned EC codes so this is used for reading GRIP and NetCDF files and then an additional software Magix which is used to actually generate the tiles that are necessary for
07:42
a web map service we have a python interface on top that we use so skinnyWMS itself is written completely in python and then uses a flask to actually deliver the WMS API and we use uwsgi to host the application skinnyWMS is available so the project home is on github
08:08
and you can easily install it by running pip install to try it out. So now back to the
08:21
features that skinnyWMS offers so as I said our aim was to provide meteorological data visualization out of the box ideally it should be interactive because usually when you have time dependent data you would like to browse through different time steps if you have
08:41
different different elevations you would like to browse through elevations and how skinny does that is basically reading the metadata of the data and then with the help of Magix actually selecting suitable visualizations and listing up available styles to create already nice
09:04
visualizations for the users. Additionally it automatically converts units where necessary for example one very common conversion is Kelvin to Celsius usually people are not very familiar if you display a temperature like 280 degrees so this is already built in and we have multidimensional
09:29
fields so so you you see time dimension and elevation dimension also in the get capabilities document which allows you then to select a specific specific data. We provide a dynamic
09:47
legend graphics which are of course very useful for for visualizations so you can select an appropriate size for your legend and then Magix and skinnyWMS together try to create the a suitable legend graphic for these dimensions and as I said we have an
10:07
interactive map viewer so once you fire up skinnyWMS pointed to the data you can open your browser and then view the data. skinnyWMS supports GRIP and NetCDF and very recently
10:23
during one of our Vismet hack hackathons this year in June we tried to add support for server side GeoJSON. GeoJSON specifically for point data station data with meteorological
10:42
parameters it is so far not very well standardized how to actually map let's say time dependent data in GeoJSON so that's that's why the support is still experimental but we are working hard on on making something nice for the meteorological community and although skinnyWMS is still quite
11:07
young it is already operationally used for example two prominent examples are the Copernicus climate data store which is available online and also GeoPortal of the
11:23
German meteorological service where skinnyWMS is used as a data preview for the open data. When it comes to features since skinnyWMS is still quite a young software we are constantly adding new features improving existing ones for example one thing that we plan to implement
11:47
would be get feature info requests which are part of the OGC standard for a web web service and this would allow us for example to then click on an on an point on the map
12:03
and then get depending on which kind of data we are visualizing for example getting time series plots or vertical profile plots at this point which would be kind of nice and of course
12:21
improving the map viewer so far it is a bit limited so it allows you to browse through different times and layers but the usability can always be improved and the NetCDF support since NetCDF is less standardized in terms of meteorological data
12:42
we would like to improve the detection of parameters in units and the out-of-box visualization for data in NetCDF and I mentioned already that we added GeoJSON format which is still experimental so we would like to better document what kind of format is required
13:06
and DWD provides already a GeoJSON converter which generates data that is can be ingested by to visualize points on the map and this GeoJSON just to add a few words
13:28
the problem is usually if you have time dependent data and you have global coverage of the data it can become quite large so if you try GeoJSON is usually very
13:41
simple to visualize on a map but when you suddenly have several hundred megabytes of data this can be quite heavy on the client so you don't especially if you have mobile devices this is not really a solution so having server-side GeoJSON support can to some degree
14:04
mitigate that because you're only transferring the data that needs to be displayed but also some complicated visualizations for example station plots that you see in the lower right
14:20
corner are quite challenging actually to create if you're using something like JavaScript so so this was one of the several reasons basically why we try to add GeoJSON so as I already mentioned skinny WMS is available on GitHub together with all our other
14:44
open source software so feel free to check it out also in skinny WMS can visualize our open data so we make more and more data available online so also make sure to have a look and
15:04
of course bug reports or feature requests or code contributions are very welcome so feel free to take a look on GitHub and since I'm the first one of the first presenting
15:21
today there are some nice upcoming talks by my colleagues Milana's talk about our open data and it can be processed and visualized James and Eddie's talk about the Copernicus climate data store which I already mentioned Dimitas and Paulo talk about the flood awareness service a very nice
15:45
application and she hands talk about the data so we we have troves of valuable and very nice data and applications so please also take a look at the talks of my colleagues and with this thank you for your attention and I'm open for questions and
16:06
comments