On the accuracy of heighting from aerial photographs and maps: Implications to process modellers

A concern to all scientists engaged in modelling landform must be the accuracy with which heights can be obtained from aerial photographs and their derivatives, conventional topographic maps. Morphological information is of importance for terrain description and analysis and increasingly as a critical boundary condition for models of geomorphological processes. This paper offers some cautionary advice in the use of aerial photographs and maps to provide morphological information. Current mapping specifications are reviewed and the prospects for improved accuracy in heighting, given new equipment such as cameras equipped with forward motion compensation, are analysed in the light of recent published material. An analysis is presented of the manner in which random errors in photogrammetric observations can be propagated through a block of aerial photographs. A further analysis is made with the addition of a small systematic error of a type and magnitude occasionally encountered in practice. These analyses show that the use of published map data (digital or not) for producing landform models should be made only after careful assessment of the accuracy of those data and that some of the claims currently being made for height accuracy obtained with new photogrammetric equipment are valid only in special cases.