Background Though physical activity (PA) is beneficial for health, large proportions of children are inactive. Understanding how weather conditions influence PA will help inform effective PA strategies aimed at increasing PA during childhood. The aim of this study was to investigate temporal and seasonal variability in objective PA among school-aged children and its association with weather conditions.

Methods Cross sectional data from 8–11 year old primary school children in Cork, Ireland. Triaxial accelerometer data was collected for 7 consecutive days during school term times (October-June) and PA was defined using Signal Magnitude Vectors (SVMgs). Children who recorded 10 h or more data during waking time for each of the 7 days were included in the analysis (467 boys and 363 girls). Data on daily and hourly weather conditions were obtained from the Irish National Meteorological Service. SVMgs normalised using average levels were used in correlation and linear mixed effect regression analyses for daily data and generalised additive mixed models for hourly data which explored the relationship between PA and weather conditions allowing for the child and school characteristics.

Results There were significant differences in PA by hour of the day and month of the year (p < 0.01). Average PA level was highest in June (SVMgs = 731) and lowest in December (SVMgs = 560). Between 7 am and 9 pm on school days, average PA was highest between 7 pm–8 pm (SVMgs = 816) and lowest between 7am-8am (SVMgs = 320). Maximum daily temperature (MDT) and rainfall were significantly associated with PA while hourly temperature, current rainfall, wind speed and cloud cover were associated with hourly PA. Daily PA increased by 1.7% for boys and 1.0% for girls for each 1°C increase in MDT. The association between rainfall and PA was the same for boys and girls. Compared to days with no rainfall, the average daily PA was 7.4% lower on days with up to 3.2mm rainfall and 16.8% lower on days with more than 3.2 mm rainfall. Hourly PA increased by 2.3% for boys and 0.4% for girls for 1°C increase in temperature. Rainfall was associated with a 10.2% reduction in hourly PA.

Conclusion There are temporal and seasonal variations in PA among school-aged children during school days and weather significantly influences hourly and daily variations in PA. This suggests that strategies to promote PA should consider season/weather appropriate PA opportunities. Both the school and home environment serve as ideal locations to increase PA in children.