This document discusses using machine learning techniques to classify crop types in the Middle Rio Grande region using multi-temporal satellite imagery. Specifically, it aims to 1) classify five main crop types (cotton, pecan, onion, pepper, alfalfa) in cultivated and non-cultivated areas using random forest classification and convolutional neural networks, and 2) determine which method achieves the highest accuracy. The methodology uses 134 bands of Landsat and other satellite data as inputs to random forest and CNN models to classify crop type and cultivation status in 2017, with the goal of improving accuracy over pixel-based methods.