Power Use Patterns Differ Between Inland and Coastal Homes in Karnataka

Researchers tracked electricity use and weather at two houses in Karnataka, India. One home sits inland near Bangalore, the other lies on the coast at Mangalore. The data came from hourly smart meters and matched weather records.

The inland house showed only a weak link between weather and power draw. The coastal house, however, reacted strongly to sunshine and also to humidity, with some lag in its response. Solar radiation was a key driver for the coastal home, while inland homes relied more on temperature swings.

When they built models using only climate data, they could explain very little of the electricity pattern, especially for the inland house. Adding past weather, interaction terms, and rolling averages helped several model types predict better, though the boost depended on the model and the house. Even simple lagged variables gave a noticeable lift for some models.

The study proves that local climate details matter for predicting home energy needs and that smart, interpretable features can make forecasts more reliable. The results show that ignoring local climate can lead to poor predictions.