This dataset contains aerial images and digital surface models of 19 out of 20 hydrological subcatchments analysed in the project "Shortcut". The subcatchments mostly lie in rural areas of the Swiss midlands and the datasets were obtained by flights with an eBee drone (SenseFly) between October 2017 and August 2018. The resolutions of the aerial images and the digital surface models lie between 2cm/pixel and 5cm/pixel, depending on the catchment. The aerial images in this dataset were used for the publication "Hydraulic Shortcuts Increase the Connectivity of Arable Land Areas to Surface Waters" (Schönenberger, U. & Stamm C, 2021)". The publication, supporting information, datasets, and codes can be found on the Eawag Research Data Institutional Repository (DOI: 10.25678/0003J3).

This dataset consists of UAS flight images from three sites along an elevation and precipitation gradient within Reynolds Creek Experimental Watershed collected between June 4 and July 9, 2019. The lowest elevation site ('wbs1', 1,425 m) was vegetated by shrub steppe dominated Wyoming big sage (Artemisia tridentata ssp. wyomingensis). Vegetation at the middle elevation site ('los1', 1,680 m) was shrub steppe dominated by low sage (Artemisia arbuscula). Shrub steppe at the highest elevation site ('mbs1', 2,110 m) was dominated by mountain big sage (Artemisia tridentata ssp. vaseyana) and Utah snowberry (Symphoricarpos oreophilus utahensis). A MicaSense RedEdge 3 sensor mounted on a DJI Matrice 600 Pro UAS platform was used to collect multispectral imagery of each site. The drone was flown by a Federal Aviation Administration (FAA) Part 107 certified remote pilot between June 5 and July 9 2019. All flights were completed within two hours of solar noon. The RedEdge is a broadband multispectral sensor: blue (475nm), green (560nm), red (668nm), red edge (717nm), and near-infrared (840nm). The RedEdge sensor was radiometrically calibrated using a reflectance panel before and after each flight. A DJI Phantom 4 with the stock FC330 Red Green Blue (sRGB) camera was flown over each site to collect imagery at a finer spatial resolution to assist with training and test data for vegetation type classification.