The article provides a comprehensive set of "good practice" recommendations for designing and implementing accuracy assessments and area estimations in land change mapping. The recommendations address three major components: sampling design, response design, and analysis. Key points include: 1. **Sampling Design**: Implement a probability sampling design to achieve accuracy and area estimation objectives while considering practical constraints. Use stratified random sampling to increase sample size in rare classes and improve precision. 2. **Response Design**: Use reference data sources that provide sufficient spatial and temporal representation to accurately label each unit in the sample. Ensure the reference classification is more accurate than the map classification being evaluated. 3. **Analysis**: Conduct analysis consistent with the sampling and response designs, reporting accuracy metrics such as error matrices, overall accuracy, user's accuracy, and producer's accuracy. Quantify uncertainty using confidence intervals. 4. **Area Estimation**: Estimate the area of classes (e.g., types of change) based on the reference classification of the sample units. 5. **Uncertainty Quantification**: Evaluate variability and potential errors in the reference classification. 6. **Deviation Documentation**: Document any deviations from good practices that may affect the results. The article also emphasizes the importance of visually assessing the remote sensing product for obvious errors and concerns before proceeding with detailed assessments. It provides a detailed guide on choosing the appropriate sampling design, including considerations for stratification, cluster sampling, and systematic vs. random selection. The response design focuses on selecting spatial units and determining the reference classification, while the analysis section outlines protocols for quantifying accuracy and area estimates. An example application is provided to illustrate the recommended process, highlighting the practical application of these good practices in real-world scenarios.The article provides a comprehensive set of "good practice" recommendations for designing and implementing accuracy assessments and area estimations in land change mapping. The recommendations address three major components: sampling design, response design, and analysis. Key points include: 1. **Sampling Design**: Implement a probability sampling design to achieve accuracy and area estimation objectives while considering practical constraints. Use stratified random sampling to increase sample size in rare classes and improve precision. 2. **Response Design**: Use reference data sources that provide sufficient spatial and temporal representation to accurately label each unit in the sample. Ensure the reference classification is more accurate than the map classification being evaluated. 3. **Analysis**: Conduct analysis consistent with the sampling and response designs, reporting accuracy metrics such as error matrices, overall accuracy, user's accuracy, and producer's accuracy. Quantify uncertainty using confidence intervals. 4. **Area Estimation**: Estimate the area of classes (e.g., types of change) based on the reference classification of the sample units. 5. **Uncertainty Quantification**: Evaluate variability and potential errors in the reference classification. 6. **Deviation Documentation**: Document any deviations from good practices that may affect the results. The article also emphasizes the importance of visually assessing the remote sensing product for obvious errors and concerns before proceeding with detailed assessments. It provides a detailed guide on choosing the appropriate sampling design, including considerations for stratification, cluster sampling, and systematic vs. random selection. The response design focuses on selecting spatial units and determining the reference classification, while the analysis section outlines protocols for quantifying accuracy and area estimates. An example application is provided to illustrate the recommended process, highlighting the practical application of these good practices in real-world scenarios.