(P23-005-25) Spatiotemporal Variability of Temperature and Its Association With Food Availability in Traditional Outdoor Food Markets in Kenya

Objectives: Rising temperatures threaten food supply chains, posing food safety and loss challenges for traditional outdoor markets, vital to food access in LMICs. We examine heat exposure, heat island intensity, and food availability in five outdoor markets over two seasons in Kenya.

Methods: A market census was conducted across an urban-rural gradient in two agroecologies in August 2024 (dry season). Food availability and vendor characteristics were collected during peak sale hours (4-7 PM). A second census was conducted in mornings (8-11 AM) and evenings (4-7 PM) in November (short rains). Temperature data were obtained: 1) ground sensors collecting ambient temperature and relative humidity to calculate vendor heat exposure; 2) Landsat 9 thermal bands to derive land surface temperature (LST); 3) nearby meteorological station. Heat island intensity quantified thermal modification of markets. Regressions predicted food availability with heat exposure, adjusted for vendor characteristics, land cover, and site clusters.

Results: Vendors (n=509 dry season; n=188 short rains) sold 3.4±3.2 food groups, including vegetables (33%), fruits (20%), meats (12%), mostly from shops (34%) or tables (24%). Heat exposure was 25.8±2.4°C in dry season and 22.5±1.6°C in short rains, with evenings 4.1°C warmer than mornings. Heat exposure varied 6-11°C within sites, highest in peri-urban Machakos. Heat island intensity varied across sites. Ground temperatures were 0.5°C warmer than meteorological readings and 5°C cooler than LST. Higher heat exposure was associated with lower raw food availability (OR 0.87 [95%CI 0.78,0.97]), particularly vegetables, and higher cooked food availability (1.20 [1.10,1.32]), including animal-source foods and shelf-stables, strongest in dry season. In short rains, higher heat exposure was associated with 85% higher odds of ready-to-eat food availability in mornings and 37% lower odds in evenings.

Conclusions: Heat exposure variations within markets suggest microclimates that reduce access to nutrient-dense foods and may reduce food security. While production versus market impact is unclear, linking ground, meteorological, and satellite data helps monitor temperature in outdoor markets to target climate-smart strategies.

Funding Sources: Innovative Methods and Metrics for Agriculture and Nutrition Actions; University of South Carolina