Accurate predictions of crop yield are critical for developing effective agricultural and food policies at the regional and global scales. We evaluated a machine-learning method, Random Forests (RF), for its ability to predict crop yield responses to climate and biophysical variables at global and regional scales in wheat, maize, and potato in comparison with multiple linear regressions (MLR) serving as a benchmark. We used crop yield data from various sources and regions for model training and testing: 1) gridded global wheat grain yield, 2) maize grain yield from US counties over thirty years, and 3) potato tuber and maize silage yield from the northeastern seaboard region. RF was found highly capable of predicting crop yields and outperformed MLR benchmarks in all performance statistics that were compared. For example, the root mean square errors (RMSE) ranged between 6 and 14% of the average observed yield with RF models in all test cases whereas these values ranged from 14% to 49% for MLR models. Our results show that RF is an effective and versatile machine-learning method for crop yield predictions at regional and global scales for its high accuracy and precision, ease of use, and utility in data analysis. RF may result in a loss of accuracy when predicting the extreme ends or responses beyond the boundaries of the training data.

Lidar remote sensing has been used to survey stream channel and floodplain topography for decades. However, traditional platforms, such as aerial laser scanning (ALS) from an airplane, have limitations including flight altitude and scan angle that prevent the scanner from collecting a complete survey of the riverscape. Drone laser scanning (DLS) or unmanned aerial vehicle (UAV)-based lidar offer ways to scan riverscapes with many potential advantages over ALS. We compared point clouds and lidar data products generated with both DLS and ALS for a small gravel-bed stream, Stroubles Creek, located in Blacksburg, VA. Lidar data points were classified as ground and vegetation, and then rasterized to produce digital terrain models (DTMs) representing the topography and canopy height models (CHMs) representing the vegetation. The results highlighted that the lower-altitude, higher-resolution DLS data were more capable than ALS of providing details of the channel profile as well as detecting small vegetation on the floodplain. The greater detail gained with DLS will provide fluvial researchers with better estimates of the physical properties of riverscape topography and vegetation.

Data concerning streambank retreat (SBR) rates are important for many different engineering applications such as stream restoration and total maximum daily load (TMDL) development. However, measurement of SBR can be time-consuming and is often characterized by large measurement and interpolation errors. These errors propagate into the calculation of sediment budgets for the development of TMDLs, creating uncertainty in source partitioning and overall load estimates. We compared two techniques for measuring SBR: (1) traditional surveying with a total station and (2) terrestrial laser scanning (TLS). An 11-m streambank on Stroubles Creek in Blacksburg, Virgina was surveyed six times over a 2-year period. The average SBR along the entire bank was estimated to be −0.15 m/year with TLS and −0.18 m/year with total station surveying. The resulting differences in median SBR estimates along five distinct cross sections between each of the survey dates ranged from −0.11 to +0.06 m . This error in SBR due to total s...