Mangroves are unique inter-tidal ecosystems that provide valuable ecosystem goods and services. This dissertation investigates new methods of characterizing mangrove forests using remote sensing with implications for mapping and modeling ecosystem goods and services. Specifically, species composition, leaf area, and canopy height are investigated for mangroves in the Galapagos Islands. The Galapagos Islands serve as an interesting case study where environmental conditions are highly variable over short distances producing a wide range of mangrove composition and structure to examine. This dissertation reviews previous mangrove remote sensing studies and seeks to address missing gaps. Specifically, this research seeks to examine pixel and object-based methods for mapping mangrove species, investigate the usefulness of spectral and spatial metrics to estimate leaf area, and compare existing global digital surface models with a digital surface model extracted from new very high resolution imagery. The major findings of this research include the following: 1) greater spectral separability between true mangrove and mangrove associate species using object-based image analysis compared to pixel-based analysis, but a lack of separability between individual mangrove species, 2) the demonstrated necessity for novel machine-learning classification techniques rather than traditional clustering classification algorithms, 3) significant but weak relationships between spectral vegetation indices and leaf area, 4) moderate to strong relationships between grey-level co-occurrence matrix image texture and leaf area at the individual species level, 5) similar accuracy between a very high resolution stereo optical digital surface model a coarse resolution InSAR product to estimate canopy height with improved accuracy using a hybrid model of these two products. The results demonstrate advancements in remote sensing technology and technique, but further challenges remain before these methods can be applied to monitoring and modeling applications. Based on these results, future research should focus on emerging technologies such as hyperspectral, very high resolution InSAR, and LiDAR to characterize mangrove forest composition and structure.