Making Open Weather Data More Accessible: Extracting Seasonal Insights from Singapore Weather Station Data

Making Open Weather Data More Accessible: Extracting Seasonal Insights from Singapore Weather Station Data
讓天氣數據資料開源-洞悉新加坡天氣站的季節性資料

Event: Open UP Global Summit 2019
Date: 1 December 2019
Location: Taipei, Taiwan

Most of the world experience the four seasons, but how many seasons does a tropical country like Singapore have? We can get such insights from publicly-available weather data; however, we need to write scripts to make API requests to retrieve the data. In this talk, I will be sharing about a data-scraping tool that I have created to retrieve Singapore weather station data for a user-defined time period, and using a data-driven approach to gain a better understanding on the weather trends and seasons over the past few years.

Transcript

1. Making Open Weather Data More Accessible: Extracting Seasonal Insights from Singapore Weather Station Data Presented by: Ong Chin Hwee (@ongchinhwee) 1 December 2019
2. About me Ong Chin Hwee 王敬惠 • Data Engineer @ ST Engineering • Background in aerospace engineering + computational modelling • Loves (and contributes to) pandas • Mentor team at BigDataX, one of the Top 10 Data Communities in Singapore @ongchinhwee
3. Singapore 新加坡: 1°17'22.81"N, 103°51'0.25"E 北纬1度，经纬103度 @ongchinhwee
4. @ongchinhwee 105° 0°
5. A year is usually split into four seasons… 常說一年划分为四季。。。 @ongchinhwee
6. @ongchinhwee 春 Spring
7. @ongchinhwee 春 Spring 夏 Summer
8. @ongchinhwee 春 Spring 夏 Summer 秋 Autumn
9. @ongchinhwee 春 Spring 夏 Summer 秋 Autumn 冬 Winter
10. ...how about my home country? 。。。新加坡有多少季節? @ongchinhwee
11. Are the “four seasons” just… @ongchinhwee
12. Are the “four seasons” just… 1. Hot 熱 2. Very Hot 很熱 3. Extremely Hot 超熱 4. Rain? 雨 @ongchinhwee
14. Let’s discover seasons through data. 讓我們以數據來發掘季節。 @ongchinhwee
15. Extracting Weather Data @ongchinhwee
16. @ongchinhwee Data.gov.sg - Singapore’s Open Data Portal 新加坡開源數據平台
17. Realtime Weather Readings across Singapore Real-time API (Application Programming Interface) on Data.gov.sg (Singapore’s open data portal) 即時應用程式介面 @ongchinhwee
18. Realtime Weather Readings across Singapore Real-time API on Data.gov.sg (Singapore’s open data portal) Open government data available under the Singapore Open Data License 政府開放資料 @ongchinhwee
19. Realtime Weather Readings across Singapore Real-time API on Data.gov.sg (Singapore’s open data portal) Open government data available under the Singapore Open Data License Minute-by-minute weather station readings @ongchinhwee
20. “Let’s try to scrap weather data for a specific weather station!” @ongchinhwee
21. “Let’s try to scrap weather data for a specific weather station!” “How about we scrap month-long data from the API?” @ongchinhwee
22. @ongchinhwee
23. @ongchinhwee
24. @ongchinhwee
25. @ongchinhwee Nested JSON format!
26. @ongchinhwee
27. @ongchinhwee
28. Open Data = Available 開放資料 = 任何人都能取得 @ongchinhwee
29. Open Data = Accessible? 開放資料 = 任何人都能使用？ @ongchinhwee
30. “The data must be available as a whole and at no more than a reasonable reproduction cost...The data must also be available in a convenient and modifiable form.” - Open Knowledge Foundation @ongchinhwee
31. Open Data - Available, Readable, Accessible Available: Open data is freely available for everyone to access and use without restrictions or exorbitant costs. @ongchinhwee
32. Open Data - Available, Readable, Accessible Available: Open data is freely available for everyone to access and use without restrictions or exorbitant costs. Readable: Open data is more usable when it is shared in a machine-readable format such as JSON @ongchinhwee
33. Open Data - Available, Readable, Accessible Available: Open data is freely available for everyone to access and use without restrictions or exorbitant costs. Readable: Open data is more usable when it is shared in a machine-readable format such as JSON Accessible: Open data is only useful if it’s shared in a way that everyone can easily discover and understand. @ongchinhwee
34. If not everyone can understand how to use the available data, is the data really open? @ongchinhwee
35. @ongchinhwee
36. @ongchinhwee
37. I ended up creating my own scraping tool. @ongchinhwee
38. @ongchinhwee “Scraping Meteorological Data from Data.gov.sg APIs” Project
39. Data.gov.sg Weather Data API Scraping Scraping weather data from APIs via Python “Requests” library “Requests”: Python library for humans to send HTTP requests @ongchinhwee
40. Data.gov.sg Weather Data API Scraping Scraping weather data from APIs via Python “Requests” library Currently supported Data.gov.sg APIs: 1. Air Temperature (in degree Celsius) -- 17 Weather Stations 2. Rainfall (in mm) -- 53 Weather Stations @ongchinhwee
41. Data.gov.sg Weather Data API Scraping Scraping weather data from APIs via Python “Requests” library Currently supported Data.gov.sg APIs: 1. Air Temperature (in degree Celsius) 2. Rainfall (in mm) Scrap data for continuous time range + specific weather station @ongchinhwee
42. API Scraping Demo @ongchinhwee
43. Singapore’s Weather Station Positions (Source: Weather.gov.sg) 新加坡气象站位置 @ongchinhwee
44. Time Series Analysis of Singapore Weather Station Data Selected weather station: Changi Weather Station (ID: S24) Analysis timeframe: 2 Dec 2016 to 30 Nov 2019 (~3 years) Objective: - Extract trend and seasonality from minute-by-minute weather time series @ongchinhwee
45. Time Series Analysis Demo @ongchinhwee
46. Key Takeaways 1. Open data needs to be accessible to anyone to be useful. ◦ Putting data out on the web with open data license is usually not enough - it’s not just developers using open data! @ongchinhwee
47. Key Takeaways 1. Open data needs to be accessible to anyone to be useful. ◦ Putting data out on the web with open data license is usually not enough - it’s not just developers using open data! 2. If an open data source is not intuitive enough to use, build a tool to make the data more accessible for everyone! @ongchinhwee
48. Reach out to me! : ongchinhwee : @ongchinhwee : hweecat https://ongchinhwee.me And check out my project on: hweecat/api-scraping-nea-datasets @ongchinhwee