There is no strict definition of outlier. Real world data has variance and one point will always land further from the rest.

Both local outlier factor and isolation forest algorithms require a contamination level. Some implementations of the algorithm will automatically calculate contamination. In this case I think both algorithms will identify point(s) the points furthest away from the arch-typical data point.

Before doing any anomaly detection, consider doing some data exploration / data visualization. Plot the time series. Make some distribution plots, box plots, scatter-plots and see there are any obvious outliers.