2016 Irrigated Lands for the Portneuf Area: Machine Learning Generated

Portneuf Irrigated Lands 2016 was created for use in water budget studies in the Portneuf area. The area of interest was determined by Hydrology Section staff at IDWR, and a study boundary was given to GIS staff and used to clip the model output. The random forest (RF) model is a type of supervised machine learning algorithm requiring GIS staff to provide manually labeled training data. GIS staff also provide the RF model with several input features, typically raster datasets that help distinguish characteristics of irrigated lands. Portneuf Irrigated Lands 2016 used the following as input features: • Interpolated Harmonized Landsat and Sentinel-2 surface reflectance imagery (bands: SWIR 2, Blue, calculated NDVI, and NIR) • 10-meter digital elevation model (including slope and aspect) • Height Above Nearest Drainage (HAND) • Highly Scalable Temporal Adaptive Reflectance Fusion Model (HISTARFM) (bands: calculated NDVI, NIR) • Mapping EvapoTranspiration at high Resolution with Internalized Calibration (METRIC) For additional information on the interpolation process for Landsat and Sentinel-2 imagery, please see below. Additional datasets used only for labeling training data include IDWR-provided Active Water Rights Place of Use, the Cropland Data Layer for 2016, and the National Agriculture Imagery Program (NAIP) imagery for Idaho 2017. The accuracy of the Portneuf Irrigated Lands 2016 dataset was verified by several methods. Firstly, a validation test is done by withholding a subset of the training data to evaluate how well the model classifies unseen information. Second, GIS staff will run several iterations of the model with variations of training data, with the goal of improving classification for areas consistently misclassed. This process requires GIS staff knowledge, aided by supplementary datasets, to review the area and make decisions. Once a model iteration is determined as ‘final’, a manual mask is created to correct any remaining misclass in the dataset. For Portneuf Irrigated Lands 2016, final review and correction of the dataset was aided by comparing with Portneuf Irrigated Lands 2021 and 2023, which were created in tandem. By comparing all three years together, patterns of irrigation and misclass were more easily identified. Consistent areas of misclass for Portneuf Irrigated Lands 2016 include the Marsh Valley/Marsh Creek area, fields south of Chesterfield, and a handful of fields in the Bancroft/Lund area that appear to be dryland farmed. Misclass largely occurred in areas with active water rights, but no visible irrigation or irrigation infrastructure in the satellite and aerial imagery available. Final decisions made for these areas were conservative, relying heavily on both an active water right and clear indication of diverted water and artificial application of water to a given field. A less-conservative version was created for Portneuf 2016 and is available on request.