Land use/land cover (LULC) change occurs when humans alter the landscape, and this leads to increasing loss, fragmentation and spatial simplification of habitat. Many fields of study require monitoring of LULC change at a variety of scales. LULC change assessment is dependent upon high-quality input data, most often from remote sensing-derived products such as thematic maps. This research compares pixel- and object-based classifications of Landsat Thematic Mapper (TM) data for mapping and analysis of LULC change in the mixed land use region of eastern Ontario for the period 1995–2005. For single date thematic maps of 10 LULC classes, quantitative and visual analyses showed no significant accuracy difference between the two methods. The object-based method produced thematic maps with more uniform and meaningful LULC objects, but it suffered from absorption of small rare classes into larger objects and the incapability of spatial parameters (e.g. object shape) to contribute to class discrimination. Despite the similar map accuracies produced by the two methods, temporal change maps produced using post-classification comparison (PCC) and analysed using intensive visual analysis of errors of omission and commission revealed that the object-based maps depicted change more accurately than maximum likelihood classification (MLC)-derived change maps.

Few countries are using space-based Synthetic Aperture Radar (SAR) to operationally produce national-scale maps of their agricultural landscapes. For the past ten years, Canada has integrated C-band SAR with optical satellite data to map what crops are grown in every field, for the entire country. While the advantages of SAR are well understood, the barriers to its operational use include the lack of familiarity with SAR data by agricultural end-user agencies and the lack of a ‘blueprint’ on how to implement an operational SAR-based mapping system. This research reviewed order of operations for SAR data processing and how order choice affects processing time and classification outcomes. Additionally this research assessed the impact of speckle filtering by testing three filter types (adaptive, multi-temporal and multi-resolution) at varying window sizes for three study sites with different average field sizes. The Touzi multi-resolution filter achieved the highest overall classification accuracies for all three sites with varying window sizes, and with only a small (< 2%) difference in accuracy relative to the Gamma Maximum A Posteriori (MAP) adaptive filter which had similar window sizes across sites. As such, the assessment of order of operations for noise reduction and terrain correction was completed using the Gamma MAP adaptive filter. This research found there was no difference in classification accuracies regardless of whether noise reduction was applied before or after terrain correction. However, implementing the terrain correction as the first operation resulted in a 10 to 50% increase in processing time. This is an important consideration when designing and delivering operational systems, especially for large geographies like Canada where hundreds of SAR images are required. These findings will encourage country-wide, regional and global food monitoring initiatives to consider SAR sensors as an important source of data to operationally map agricultural production.

The water cloud model (WCM) can be inverted to estimate leaf area index (LAI) using the intensity of backscatter from synthetic aperture radar (SAR) sensors. Published studies have demonstrated that the WCM can accurately estimate LAI if the model is effectively calibrated. However, calibration of this model requires access to field measures of LAI as well as soil moisture. In contrast, machine learning (ML) algorithms can be trained to estimate LAI from satellite data, even if field moisture measures are not available. In this study, a support vector machine (SVM) was trained to estimate the LAI for corn, soybeans, rice, and wheat crops. These results were compared to LAI estimates from the WCM. To complete this comparison, in situ and satellite data were collected from seven Joint Experiment for Crop Assessment and Monitoring (JECAM) sites located in Argentina, Canada, Germany, India, Poland, Ukraine and the United States of America (U.S.A.). The models used C-Band backscatter intensity for two polarizations (like-polarization (VV) and cross-polarization (VH)) acquired by the RADARSAT-2 and Sentinel-1 SAR satellites. Both the WCM and SVM models performed well in estimating the LAI of corn. For the SVM, the correlation (R) between estimated LAI for corn and LAI measured in situ was reported as 0.93, with a root mean square error (RMSE) of 0.64 ${\mathrm{m}}^2{\mathrm{m}}^{-2}$ and mean absolute error (MAE) of 0.51 ${\mathrm{m}}^2{\mathrm{m}}^{-2}$. The WCM produced an R-value of 0.89, with only slightly higher errors (RMSE of 0.75 ${\mathrm{m}}^2{\mathrm{m}}^{-2}$ and MAE of 0.61 ${\mathrm{m}}^2{\mathrm{m}}^{-2}$) when estimating corn LAI. For rice, only the SVM model was tested, given the lack of soil moisture measures for this crop. In this case, both high correlations and low errors were observed in estimating the LAI of rice using SVM (R of 0.96, RMSE of 0.41 ${\mathrm{m}}^2{\mathrm{m}}^{-2}$ and MAE of 0.30 ${\mathrm{m}}^2{\mathrm{m}}^{-2}$). However, the results demonstrated that when the calibration points were limited (in this case for soybeans), the WCM outperformed the SVM model. This study demonstrates the importance of testing different modeling approaches over diverse agro-ecosystems to increase confidence in model performance.

Sentinel-1 and Sentinel-2 satellites are delivering data at the spatial resolutions and geographies needed for operational crop monitoring. Of particular interest, Sentinel-1 Synthetic Aperture Radar (SAR) data in Single Look Complex (SLC) format can be processed to provide polarimetric parameters sensitive to crop development. This study created a Sentinel-1 based SAR vegetation index for canola, calibrated to Normalized Difference Vegetation Index (NDVI) (SAR_cal-NDVI) values calculated from Sentinel-2. A Random Forest Regressor (RFR) modelled the SAR_cal-NDVI, from four selected polarimetric features (degree of linear polarization (DoLP), normalized Shannon entropy (SE), the second eigenvalue of the coherency matrix (l2) and the ellipticity angle ( $\chi ))$. The coefficient of determination (R²) between the Sentinel-2 NDVI and the SAR_cal-NDVI was 0.89. RFR-generated SAR_cal-NDVI estimates, based on the four SLC generated polarimetric parameters, were then used with a Canopy Structure Dynamics Model (CSDM) and with Growing Degree Days (GDD) to estimate the condition of canola at a daily time step. The SAR_cal--NDVI time series estimated from the CSDM model was correlated to the ground measured biomass. During the rapid accumulation of biomass from early to mid-season, correlations of the SAR_cal-NDVI to wet biomass were strong (R² of 0.88). Correlations were still significant albeit weaker during pod development and throughout the period of canola senescence (R² of 0.42). As climate variability drives uncertainty in the agricultural sector, sensors like the Sentinels can be leveraged to track changes in crop acreages and crop productivity. The next step in this research is to extend to other economically important crops such as corn, soybeans and wheat and test the ability of a Sentinel-1-based vegetation index to inform crop yield estimates.