Estimating the depth of buried hot features from thermal IR remote sensing data: a conceptual approach