Aerial surveys are most effective on relatively flat ground with sparse vegetation. They cannot be used to survey forest elephants, which are difficult to see from the air. Aerial surveys require significant resources including an airplane and specialized equipment, as well as a pilot and experienced observers.
Sample counts assume that the count performed in the sampling area can be extrapolated to the entire survey area. Nonuniform elephant distributions can lead to undercounting or overcounting, depending on the herd distribution at the time of the survey. Undercounting can also occur when calves hide under an adult's belly or in the center of the herd. Overcounting can occur when elephants that have already been counted move to an uncounted transect and are counted again. Total counts can be more accurate but are too expensive and time-consuming to undertake in a large survey area.
Population estimates based on aerial surveys can be imprecise due to variations in the landscape and elephant distribution, and human sampling error. However, researchers can reduce errors by selecting the appropriate type of aerial survey for the landscape and elephant distribution, by keeping flight speed and height consistent, and by double-checking observers' identifications.