Two empirical simple linear regression models were developed from SedCam imagery and concurrent physical sediment samples over a 20-month period at the East Branch Brandywine Creek gage (USGS 01480870). The image files included here are a subset, used in the calibration dataset for these regression models. Models relate the explanatory variable, Rmax (maximum digital number of the red band, which peaks at a wavelength of 590 nanometers) to point and cross-sectional mean sample response variables (milligrams per liter). These regression models predict suspended-sediment concentration from a non-contact sediment surrogate technique to improve surface water monitoring operations. Files include Nikon Electronic Format (.NEF) raw files, Adobe Photoshop sidecar (.xmp) files, and post-processed (.tif) files for ten frames.