Measuring gullies by synergetic application of UAV and close range photogrammetry — A case study from Andalusia, Spain

• Methodological case study to assess gully headcut changes in Andalusia
• Integration of terrestrial images can improve 3D gully model of UAV images.
• Fused high resolution gully model enables analysis of 3D volume changes.

Gullies naturally exhibit complex morphologies and hence are difficult to measure. Photogrammetric methods are promising tools to overcome such problems due to 3D reconstruction from overlapping images without disturbing the surface. Airborne (e.g. unmanned aerial vehicles, UAV) and terrestrial image acquisition platforms are possible data sources for comprehensive digital gully modelling. In this study, an approach for the synergetic utilisation of UAV and terrestrial images for high resolution 3D model generation is introduced. While proper UAV mission planning is necessary, terrestrial images can be acquired spontaneously and flexibly depending on the encountered terrain conditions to supplement UAV data. Without confining accuracy demands, terrestrial images are processed solely with additional information from already existent 3D models and orthophotos from UAV image processing. During this field campaign UAVs are suitable for flexible and frequent gully monitoring. However, with the case specific flight configuration, 3D modelling is restricted at steep gully walls – above slope gradients of 50–60° – as well as overhanging gully walls, due to the limiting birds-eye view. Additional terrestrial images are used to overcome these limitations and fill up data gaps. In this study the final merged digital gully model reveals a resolution about 0.5 cm and an accuracy of 1 cm. A gully headcut in Andalusia (Spain) is analysed to confirm the suitability of the synergetic data utilisation. A study period of three months during winter season is investigated to measure multi-temporal 3D volume changes. The positive volume balance amounts to 5 m3 for a gully area of 170 m2 and is caused mainly by anthropogenic interactions. Concluding, the combination of UAV and terrestrial data allow for comprehensive gully models with high spatial resolution at frequent intervals.