Assessing the effect of weather on human outdoor perception using Twitter

  • Laura GiuffridaEmail author
  • Hanna Lokys
  • Otto Klemm
Special Issue: Subjective approaches to thermal perception


Human comfort in outdoor spaces (HCOS) is linked to people’s psychological responses to environmental variables. Previous studies have established comfort ranges for these variables through interviews and questionnaires, reaching only limited populations. However, larger amounts of data would not only generate more robust results in local studies, but it would also allow for the possibility of creating an approach that could be applied to a wider range of weather conditions and different climates. Therefore, this study describes a new methodology to assess people’s perception of weather based on human responses to weather conditions extracted from tweets, with the purpose of establishing comfort ranges for environmental variables. Tweets containing weather-associated keywords were collected, stored, and then linked to real-time meteorological data acquired nearby the locations in which the tweets were posted. Afterwards, people’s perception of weather was extracted from the tweets using a classifier trained specifically on weather data that identified irrelevant, neutral, positive, and negative tweets. The obtained tweets and their related atmospheric data were analyzed to establish comfort ranges. The tweets’ responses to effective temperature were very similar to those obtained in previous studies, although the peak of comfort is shifted towards the cold stress. Similarly, the tweets’ responses to the thermohygrometric index were alike to previous results, but the peak of comfort is shifted towards the heat stress. Regarding the single weather variables under study, the obtained comfort ranges are similar to the ones found in previous research; in particular, the temperature comfort range matches perfectly at 20–22 °C. Therefore, it was concluded that tweets can be used to assess HCOS; not only are the results of this methodology comparable to results obtained in previous studies, but the procedure itself also shows new features and unexpected future applications.


Human comfort Perception analysis Natural language classifier Sentiment analysis Weather perception Tweet analysis 



The Twitter API, the OpenWeatherMap API, the Watson Natural Language Classifier API, and the Crowdflower platform are gratefully acknowledged for providing the data and tools employed in this analysis. We thank C. Brennecka for language editing of the final version of the manuscript.


  1. Ahmed W (2015a) Using Twitter as a data source: an overview of current social media research tools. Accessed 14 January 2017
  2. Ahmed W (2015b) Challenges of using Twitter as a data source: an overview of current resources. Accessed 14 January 2017
  3. Asanet (2012) Social science with social media. Accessed 14 January 2017
  4. Austin BJ (2014) Perspectives of weather and sensitivities to heat: social media applications for cultural climatology. Kent State UniversityGoogle Scholar
  5. Bekafigo MA, McBride A (2013) Who tweets about politics? Political participation of Twitter users during the 2011 gubernatorial elections. Soc Sci Comput Rev 31(5):625–643CrossRefGoogle Scholar
  6. Biever C (2010) Twitter mood maps reveal emotional states of America. New Scientist 207(2771):14CrossRefGoogle Scholar
  7. Blazejczyk K (1994) New climatological- and -physiological model of the human heat balance outdoor (MENEX) and its applications in bioclimatological studies in different scales. Zeszyty IGiPZ PAN 28:27–58Google Scholar
  8. Blazejczyk K, Epstein Y, Jendritzky G, Staiger H, Tinz B (2012) Comparison of UTCI to selected thermal indices. Int J Biometeorol 56(3):515–535CrossRefGoogle Scholar
  9. Bollen J, Mao H, Pepe A (2011) Modeling public mood and emotion: Twitter sentiment and socio-economic phenomena. ICWSM 11:450–453Google Scholar
  10. Cox J, & Plale B (2011) Improving automatic weather observations with the public Twitter stream. Dissertation, IU School of Informatics and ComputingGoogle Scholar
  11. Crowdflower (2017) Accessed 23 January 2017
  12. de Montigny L, Ling R, Zacharias J (2011) The effects of weather on walking rates in nine cities. Environ and Behav 44(6):821–840CrossRefGoogle Scholar
  13. Fanger PO (1972) Thermal comfort: Anal Appl Environ Eng. Mc. Graww Hill, New YorkGoogle Scholar
  14. Feldman R (2013) Techniques and applications for sentiment analysis. Commun ACM 56(4):82–89CrossRefGoogle Scholar
  15. Gagge AP, Fobelets AP, Berglund PE (1986) A standard predictive index of human response to the thermal environment. ASHRAE Trans 92:709–731Google Scholar
  16. Gagge AP, Stolwijk J, Nishi Y (1971) An effective temperature scale based on a simple model of human physiological regulatory response. ASHRAE Trans 77(1):247–262Google Scholar
  17. Gagge AP, Stolwijk JAJ, Hardy JD (1967) Comfort and thermal sensations and associated physiological responses at various ambient temperatures. Enviro Res 1:1–20CrossRefGoogle Scholar
  18. Gagge AP, Stolwijk JAJ, Saltin B (1969) Comfort and thermal sensations and associated physiological responses during exercise at various ambient temperatures. Enviro Res 2(3):209–229CrossRefGoogle Scholar
  19. Gerbaudo P (2012) Tweets and the streets: social media and contemporary activism. Pluto PressGoogle Scholar
  20. Go A, Bhayani R, & Huang L (2009) Twitter sentiment classification using distant supervision. CS224N Project Report, Stanford, 1(12)Google Scholar
  21. Greenwood S, Perrin A, & Duggan M (2016) Social media update 2016. Accessed 16 January 2017
  22. Hannak A, Anderson E, Barrett LF, Lehmann S, Mislove A, & Riedewald M (2012) Tweetin’in the rain: exploring societal-scale effects of weather on mood. In ICWSMGoogle Scholar
  23. High R. (2012) The era of cognitive systems: an inside look at IBM Watson and how it works. IBM Corporation, RedbooksGoogle Scholar
  24. Höppe P (1984) Die Energiebilanz des Menschen. Wiss Mittl Meteorol Inst Uni München 49Google Scholar
  25. Höppe P (1999) The physiological equivalent temperature—a universal index for the biometeorological assessment of the thermal environment. Int J Biometeorol 43:71–75CrossRefGoogle Scholar
  26. Horton RG (2014) Ch16: Northeast. Climate change impacts in the United States: the third national climate assessment. In: Melillo JM, Richmond TC, Yohe GW, (eds) US global change research program, p 371–395Google Scholar
  27. Jung J, Uejio CK (2017) Social media responses to heat waves. Int J Biometeorol:1–14Google Scholar
  28. Katsuura T, Tachibana ME, Okada A, Kikuchi Y (1993) Comparison of thermoregulatory responses to heat between Japanese Brazilians and Japanese. J Therm Biol 18(5–6):299–302CrossRefGoogle Scholar
  29. Köppen W (1900) Versuch einer Klassifikation der Klimate, vorzugsweise nach ihren Beziehungen zur Pflanzenwelt. – Geogr. Zeitschr, 6, 593–611, 657–679Google Scholar
  30. Kottek M, Grieser J, Beck C, Rudolf B, Rubel F (2006) World map of the Köppen-Geiger climate classification updated. Meteorol Zeitschrift 15(3):259–263CrossRefGoogle Scholar
  31. Kouloumpis E, Wilson T, Moore JD (2011) Twitter sentiment analysis: the good the bad and the omg! Icwsm 11(538–541):164Google Scholar
  32. Krieck M, Dreesman J, Otrusina L, & Denecke K (2011) A new age of public health: identifying disease outbreaks by analyzing tweets. In Proceedings of Health Web-Science Workshop, ACM Web Science ConferenceGoogle Scholar
  33. Kumari P, Singh S, More D, Talpade D, Pathak M (2015) Sentiment analysis of tweets. Int J of Sci Technol & Eng 1(10):130–134Google Scholar
  34. Lachlan KA, Spence PR, Lin X, Najarian KM, Greco MD (2014) Twitter use during a weather event: comparing content associated with localized and nonlocalized hashtags. Commun Stud 65(5):519–534CrossRefGoogle Scholar
  35. Masterson J, Richardson FA (1979) Humidex: a method of quantifying human discomfort due to excessive heat and humidity. Environ Can, Downsview, OntarioGoogle Scholar
  36. McCormick TH, Lee H, Cesare N, Shojaie A, Spiro ES (2015) Using Twitter for demographic and social science research tools for data collection and processing. Sociol Methods & Res 46(3):390–421CrossRefGoogle Scholar
  37. Mehta V (2007) Lively streets determining environmental characteristics to support social behavior. J of Plan Education and Res 27(2):165–187CrossRefGoogle Scholar
  38. Missenard FA (1933) Température effective d’une atmosphere Généralisation température résultante d’un milieu. In: Encyclopédie Industrielle et Commerciale, Etude physiologique et technique de la ventilation. Librerie de l’Enseignement Technique, Paris, pp 131–185Google Scholar
  39. Nakov P, Rosenthal S, Kiritchenko S, Mohammad SM, Kozareva Z, Ritter A, Stoyanov V, Zhu X (2016) Developing a successful SemEval task in sentiment analysis of Twitter and other social media texts. Lang Resour and Eval 50(1):35–65CrossRefGoogle Scholar
  40. Nikolopoulou M, Lykoudis S (2006) Thermal comfort in outdoor urban spaces: analysis across different European countries. Build Environ 41(11):1455–1470CrossRefGoogle Scholar
  41. Nikolopoulou M, Steemers K (2003) Thermal comfort and psychological adaptation as a guide for designing urban spaces. Energy and Buildings 35(1):95–101CrossRefGoogle Scholar
  42. NRCC (2017) Accessed 19 January 2017
  43. NOAA (2017) Accessed 19 January 2017
  44. Ogunsote OO, Prucnal-Ogunsote B (2002) Comfort limits for the effective temperature index in the tropics: a Nigerian case study. Architectural Sci Rev 45(2):125–132CrossRefGoogle Scholar
  45. OpenWeatherMap (2012) Accessed 16 January 2017
  46. Pak, A., & Paroubek, P. (2010, May) Twitter as a corpus for sentiment analysis and opinion mining. In LREc (Vol. 10, No. 2010)Google Scholar
  47. Palutikof JP, Agnew MD, Hoar MR (2004) Public perceptions of unusually warm weather in the UK: impacts, responses and adaptations. Clim Res 26(1):43–59CrossRefGoogle Scholar
  48. Paul MJ, Dredze M (2011) You are what you Tweet: analyzing Twitter for public health. ICWSM 20:265–272Google Scholar
  49. Resch B, Summa A, Sagl G, Zeile P, & Exner JP (2015) Urban emotions—geo-semantic emotion extraction from technical sensors, human sensors and crowdsourced data. In Progress in location-based services, 199–212Google Scholar
  50. Rothfusz LP (1990) The heat index equation. NWS Southern Region Technical Attachment, SR/SSD 90–23, Fort Worth, TexasGoogle Scholar
  51. Sasaki R, Yamada M, Uematsu Y, Saeki H (2000) Comfort environment assessment based on bodily sensation in open air: relationship between comfort sensation and meteorological factors. J of Wind Eng and Industrial Aerodyn 87(1):93–110CrossRefGoogle Scholar
  52. Skoric M, Poor N, Achananuparp P, Lim EP, Jiang J (2012) Tweets and votes: a study of the 2011 Singapore general election. In System Science (HICSS), 45th Hawaii International Conference:2583–2591Google Scholar
  53. Stathopoulos T, Wu H, Zacharias J (2004) Outdoor human comfort in an urban climate. Build Environ 39(3):297–305CrossRefGoogle Scholar
  54. Thom EC (1959) The discomfort index. Weatherwise 12:57–60CrossRefGoogle Scholar
  55. Thorsson S, Honjo T, Lindberg F, Eliasson I, Lim EM (2007) Thermal comfort and outdoor activity in Japanese urban public places. Environ and Behav 39(5):660–684CrossRefGoogle Scholar
  56. Tumasjan A, Sprenger TO, Sandner PG, Welpe IM (2010) Predicting elections with Twitter: what 140 characters reveal about political sentiment. ICWSM 10:178–185Google Scholar
  57. Walton D, Dravitzki V, Donn M (2007) The relative influence of wind, sunlight and temperature on user comfort in urban outdoor spaces. Build Environ 42(9):3166–3175CrossRefGoogle Scholar
  58. Yaglou CP, Minard D (1957) Control of heat casualties at military training centers. Am Med Assoc Arch Ind Health 16:302–316Google Scholar

Copyright information

© ISB 2018

Authors and Affiliations

  1. 1.Institute of Landscape Ecology – Climatology GroupWestfälische Wilhelms-Universität MünsterMünsterGermany

Personalised recommendations