Data‐driven predictions of summertime visits to lakes across 17 <scp>US</scp> states

Using a dataset of more than 51,000 US lakes, we estimated the relationship between summertime lake visits, water clarity, landscape features, and other amenities, where visits were with counts geo-located photographs. Given size complexity our dataset, used combination machine learning techniques, imputation Poisson count model to estimate these relationships. We found that every additional meter average Secchi depth was associated at least 7% all else equal. Second, such as beaches, boat launches, public toilets, powerful predictors clarity. Third, strongly influenced by lake's accessibility its distance nearby lakes amenities offered. Our research highlights need for (1) better understanding how representative social media data are actual recreational behavior, (2) development best practices account nonrandom patterns in missing natural feature data, (3) potential endogeneity visit–water quality