 Hi, my name is Miguel Trebley. I am a physicist working at the Government of Canada. My co-authors are Alouan Fan, the Director of the Board of Wikimedia Canada, and Pierre Chauvet, the developer for this project. I am speaking to you today from Montreal Connect. I'm glad to take a few minutes with you to expose how a combination of Wikimedia Commons, Wikidata, and Wikipedia can be used to disseminate government authoritative data, and especially to compute over this data, to share algorithms and to share templates. In Canada, the structural charge of the weather forecast is called the Meteorological Service of Canada. It is under the Environment Federal Department, Untitled Environment and Climate Change Canada. This is where I work. In 2019, a two-year funding was accorded by the Meteorological Service of Canada to Wikimedia Canada, the local chapter of the Wikimedia Foundation, to upload our weather station observations into the different Wikimedia projects. What are weather observations? Well, in order to make a weather forecast, you need observations, most of the time, coming from instruments like those shown in this picture. They measure temperatures, atmospheric pressure, precipitation amount, et cetera. The devices are grouped in one location, which is called a weather station. Typical time intervals for reports from weather stations are hourly, daily, and monthly. It is also possible to make a summary since the station exists for an example for record values. This is called the Nalmenac Summary. For this project, we decided to upload only the monthly Nalmenac report in comments. Here is a map of all weather stations that once existed or still exist in Canada. There are a lot of them because there have been observations for a long time in our country. The oldest observation dates from 1840 in Toronto and many stations have been opened and closed since then. There are more than 8,000 stations times two per 10 per stations, 16,000 reports, that gives 16,000 reports in the dataset. Now that we know a little bit about the input dataset, let's go through the list of the Wikimedia projects that we use to manipulate this data. The first one is Wikimedia Commons, the media repository, accessible to all Wikimedia projects. In late 2016, an additional capability was added to Commons, the capacity to upload data in GCEN format. It didn't seem to cause much of a stir at the time and even now, but it was this additional capacity that allowed us to upload the Canadian weather observations in Commons. This is what the data file looks like in Commons. It is an example of weather data that can be found. A second tool that we used is Wikidata. We use it to upload the station metadata. We put the information of every station's location, data opening, closing, type of observations, et cetera, in every station element we created. We then link Wikidata station element with their corresponding observation files in Commons. Doing that, we now have one point of entry containing the station metadata and the path to retrieve the corresponding observations in Commons. And finally, we use Wikimedia in French to display the data using a template. In this example, an element file located in Commons is sparse to extract record values for both temperatures and precipitation for every month of the year. The table is updated when the data source on Commons is changed. In other words, we have a dynamic table on Wikimedia linked to the data source in Commons. Several tools have been developed to convert and manipulate the original data, not only to import it in Commons and Wikidata. Since this is lightning talk, I'll go through only a few of them. An interactive map was created to select a subset of weather stations. In this example, I selected the stations of the region where I was born, Sagnilax Engine. I used another tool to combine the Almanac built-in of these stations and then uploaded the result in Commons. This combination of station gives a climate overview of the region. This is what you've seen earlier in the Wikipedia page, that table. Finally, we wrote sparkle requests to identify weather stations using Wikidata. As a specific example, showing the power of combining data already in Wikidata, you can see on this map a result of a sparkle request identifying all the weather stations in Quebec that are within one kilometer of a bridge that is longer than 100 meters. This could be used, for example, to identify extreme weather values for which a bridge should be built to endure. In January 2020, Wikimedia Canada held a brainstorming session with participants from various backgrounds, academic and geos, open data movement, government, et cetera. More than 60 people participated, which, well, surprised us a bit, to be honest. There is a clear interest in how Media Wiki Foundation projects could be used to disseminate and manipulate data of this type. Several ideas for data reuse were submitted. I'll name a few here to what you're appetite. First idea, on an interactive graph, a user could choose a start and an end date to see the evolution of the climate over this period in a specific location. These graphs could be linked to the demographic corresponding to that location. This would show how many people in a given area are affected by changes in precipitation, heat, et cetera. By adding geolocation, someone logging into Wikimedia could see, depending on the area where he or she is, a personalized impact of climate change. This would allow one to see the changes at a very local level, rather than thinking that it is far away and only happening to other people. Second idea, you are probably familiar with the info box that we find on Wikimedia articles. For any info box containing the location and the date, we could add the weather observations of the closest weather station. That could be useful to describe some events affected by weather condition, like forest fire. Third idea, all photos taken in Canada with the latitude, the longitude location and date could be linked to the closest weather observations in observation in space and time. On the reverse, we could also have a bank of commons photos corresponding to a certain weather condition. One could, for example, retrieve all the pictures of a very wet period over Canada for 2017. This feature could also be useful for weekly travel. Conclusion, there are many reasons why someone would want to import data in Wikimedia. The main reason is to make these data accessible, reusable and publishable without any restriction alongside with all the other information already available. This is what is exciting about this project, in my opinion. We are expending the scope of knowledge of the Wikimedia Foundation project by including Radata. From the moment the user has a web browser, he or she can access the data and manipulate it. Next step for us. Well, the first step is the simply S. It is to translate the template to display the record values shown earlier in other Wikimedia, starting with Wikimedia in English. We hope in this way to increase interest in the use of the data. We would also like to add data from other countries. For licensing reasons, it would be easy for us to upload the US weather observation to Cyclo. Three, we were recently contacted by the World Meteorological Organization, United Nations Agency. They are in developing new standard for disseminating observations to every countries and they are interested in exploring with us the possibility of aligning this project, described in this talk with their mandate. And four, finally, the presentation we have made so far have generated some interest. This summer, two data journalism interns are publishing article addressing climate change issues in Canada, guided by the data we've put in comments in Wikidata. We are also looking to expand the team of participants and people involved in weather data ingestion. If you are interested, you can contact us by this project or by joining, contributing. You can contact us by email or through any of our Wiki user page. Moreover, if you know people in your corresponding national weather organization, please tell them about this project. If you need more information, you can contact me or other one fan at Discord.net or by joining us at our unconference stable B at 13 UTC today in 30 minutes. Thank you.