Image Change Detection (ICD) methods are widely adopted to update large area land use/cover products. Uncertainty problems, however, are well known in such techniques, and a transparent assessment is necessary. In this study, a framework was proposed for evaluating binary land change utilizing remote sensing images. First, two widely adopted ICD methods were used to establish change maps. Second, binary decisions on Change (C) and Non-Change (NC) classes were reached through thresholding on change maps. Then, results were evaluated using two sampling designs: random sampling and stratified sampling. Analysis of results suggests that (1) for random sampling, with an increasing threshold on change variables, the overall accuracy increases and shows a large variance, which is highly correlated with the C omission error; and (2) comparatively, for stratified sampling, in which two strata (i.e., C and NC) were set, the overall accuracy shows a smaller variance and is highly associated with the NC commission error. The significant trends in accuracy assessments indicate the trade-offs between the C and NC classification errors in a binary decision and can present superficial or perfunctory accuracy evaluation in certain circumstances that the causes of error sources and uncertainty problems in ICD are not fully understood.Techniques for Image Change Detection (ICD), that automatically correlate and compare sets of images taken from the same area at different times [9], are widely adopted to automate or semi-automate the updating of LUCC products. Over the past decades, numerous ICD methods, using geospatial data, have been developed [9][10][11][12][13][14][15][16]. In this study, we only focused on bi-temporal (before-after) and binary (Change (C)/Non-Change (NC)) land change detection. To facilitate our work, we divided ICD techniques into two categories: bi-temporal image analysis and image-and-map analysis, based on whether thematic land classes are used to facilitate the image analysis or not. Bi-temporal image analysis is favorable for its high efficiency and effectiveness, by implementing analysis at a pixel level without ground truth data [11]. Nevertheless, Canty [17] pointed out that if change detection is carried out at the pixel level (as opposed to using segments or objects), error (typically >5%) may corrupt or even dominate the true change signal depending on its strength. Instead of adopting bi-temporal images, "image-and-map analysis" takes the thematic maps as auxiliary data in ICD. As images are not directly compared to each other during the land change detection process, radiometric and phenology differences could be well resolved [18]. The "Cross-Correlation" algorithm, designed by Koeln and Bissonnette [19], can be used to evaluate the differences between an existing land cover map (Date 1) and a recent single-date multispectral image (Date 2) [11,16,20]. Unlike bi-temporal image analysis, pixel-based spectral analysis can contribute to spatial analysis at an object-level by segmenting heterogeneou...