6.4 Media Analysis: Communication of the 2015/16 El Niño in Kenya

Wednesday, 25 January 2017: 9:15 AM
612 (Washington State Convention Center )
Elisabeth Gawthrop, IRI, Palisades, NY; and D. J. Farnham, F. Fiondella, A. S. Owusu, M. Thomson, M. Ambani, F. Percy, W. Chung, and K. McKeown

The 2015/2016 El Niño has provided an unprecedented opportunity to analyze the dissemination of specific climate event information from a range of providers, its potential, and realized impacts. In this effort, the International Research Institute for Climate and Society (IRI) has partnered with the Columbia University Data Science Institute (DSI) and CARE/ALP of Kenya to evaluate El Niño references in social media and local media analysis (i.e. media’s use Swahili, framing of stories, use of radio and other local media), respectively. The DSI is currently incorporating machine learning into a custom-made Perceptron (Decision Tree) to predict the number of El Niño Likes/Comments (or other combination of these indicators) on Facebook. These results will ultimately indicate which news articles and sources Kenyans prefer for El Niño related information, sentiment, and if any content is linked to the Kenya National Meteorological and Hydrological Services (NMHS). Kenyan local El Niño media analysis is currently being contracted by IPSOS, a global market research firm, to extract data within four key areas of Kenyan television, radio, and print (both paper and online):
  1. Preferred media channels and languages, specifically the frequency of El Niño media communications at different levels and locations from different media channels over time (before, during, and after the El Nino)
  2. Sources of El Niño information used in stories, such as the NMHS, Red Cross Societies or the private sector
  3. Sentiment and framing of media coverage, tracking, and transformation over time and in relation to different climate information sources, climate impacts, medium, issues, sectors, responses, locations etc.
  4. Predictions versus actual El Niño 2015/16, inclusive of (1) relative coverage of 2015 El Niño forecasts, (2) past El Niño comparison and (3) potential outcomes and impacts (including speculation) versus actual impacts

Additionally, CARE/ALP impact assessments have been expanded to explicitly include assessment of El Niño communication for the October-November-December 2015 season, comparing advisories and actions of OND 2015 and the preparation and actions for March-April-May 2016. Aimed at improving the approach for communicating climate information and uncertainty to different users, the CARE/ALP impact assessments also evaluated 4 seasons of Kenya NMHS advisories in all 47 counties of Kenya assessing the quality of the advisories developed over time, their reach and use with gender and diverse user considerations (e.g. farmers, pastoralists, county agriculture and livestock staff, County Directors of NMHS, etc.).

** Please note that this analysis (social media, local media, and impact assessments) is presently concluding via the support of DFiD WISER ENACTS. A range of traditional, illustrative and interactive graphics are currently being created to show how Kenyans communicated and interpreted the 2015/16 El Niño based on the data collected. Thus, this presentation intends to creatively demonstrate the communication of weather and climate information through national and local media in Kenya.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner