2024-11-10-世界银行-国家财富的变化——非木材森林生态系统服务经济价值的全球评估_技术报告（英）_220页_6mb

Summary of The Changing Wealth of Nations Technical Report: Global Assessment of the Economic Value of Non-Wood Forest Ecosystem Services

Core Content

This report updates and extends the methodology from the previous Changing Wealth of Nations (CWON 2021) to assess the economic value of non-wood forest ecosystem services (NWFS) globally. It provides a comprehensive analysis of the valuation of these services, including recreation, hunting, and fishing; non-wood forest products; and water services, and evaluates the contribution of protected areas to these values.

Main Points

• Objective: To estimate the economic value of non-wood forest ecosystem services at a global scale using a spatially explicit approach.
• Scope: The report focuses on four types of non-wood forest ecosystem services: recreation, hunting, and fishing; non-wood forest products (NWFP); water services; and habitat and species protection.
• Methodology: A meta-analytic predictive model is developed using regression and machine learning techniques to estimate the value of these services across different regions and ecosystems.
• Database Development: A global database of primary value estimates is constructed, incorporating new studies and updating previous data.
• Spatial Resolution: The analysis is conducted at a 0.1° by 0.1° (approximately 10km by 10km) grid resolution, providing local-level estimates.
• Protected Areas Contribution: An operational method is introduced to assess the contribution of protected areas to the value of non-wood forest ecosystem services.

Key Information

1. Database Development

• The database was updated by incorporating 79 new primary value estimates from 677 additional studies, mostly published after the previous database was developed.
• The largest increases in new value estimates were from Asia (34%), Africa (32%), and Europe (19%).
• The report includes 28 new estimates for recreation, 20 for habitat and species protection, 17 for non-wood forest products, and 14 for water services.
• The database includes studies from various sources, such as:
  • TEEB, EVRI, and ESDV.
  • Peer-reviewed journals like Environmental and Development Economics, Journal of Environmental Economics and Management, Science, and Nature.
  • Grey literature, including reports, books, theses, and non-journal publications.

2. Valuation Methods

• The report uses meta-regression analysis and machine learning (ML) techniques to estimate the value of ecosystem services.
• ML models include:
  • Random Forest (RF)
  • Bayesian Additive Regression Trees (BART)
  • Automated Machine Learning (AML)
• Variable selection is conducted using VSURF (Variable selection using random forest) to identify the most important predictors.
• The report also discusses model comparison and statistical criteria for model selection, such as RMSE (Root Mean Square Error) and AIC/BIC (Akaike/Bayesian Information Criteria).

3. Geographical and Ecosystem Service Analysis

• Geographical distribution of value estimates is analyzed, with the highest values observed in Europe, the United States, the Republic of Korea, Japan, and Australia.
• Recreation services have the highest average marginal value at US$48/ha/year, while non-wood forest products have the lowest at US$6.80/ha/year.
• Protected areas contribute 16% to the total value of recreation, hunting, fishing, non-wood forest products, and water services globally in 2020.
• Habitat and species protection values are not included in the combined value of non-wood forest ecosystem services due to their inclusion in SEEA EA (System of Environmental-Economic Accounting - Ecosystem Accounting), which excludes non-use values.

4. Global Predictions and Regional Analysis

• Global predictions are generated for each ecosystem service, using the world grid and country-level aggregations.
• The results show substantial heterogeneity in the values of non-wood forest ecosystem services across different regions and services.
• Regional and country-level average predictions are provided for 2020, with the total wealth of non-wood forest ecosystem services calculated in constant USD 2020.

5. Methodological Extensions

• The methodology is extended by:
  • Using revised and additional machine learning approaches, including BART.
  • Adding a section on the alignment of valuation literature with SEEA EA.
  • Developing an operational method to assess the contribution of protected areas to the value of non-wood forest ecosystem services.

Structure of the Report

• Section 1: Introduction – outlines the importance of non-wood forest ecosystem services and the objective of the report.
• Section 2: Database development – includes a literature review, data extraction, and update of the valuation database.
• Section 3: Methods for spatial estimation – discusses the use of regression and machine learning techniques for predicting ecosystem service values.
• Section 4: Database construction and descriptive analysis – describes the process of linking the world grid to study sites and provides descriptive statistics.
• Section 5: Estimation results and global predictions – presents the results of the meta-regression and ML models, and the global and country-level predictions.
• Section 6: Contribution of protected areas – introduces the operational method for assessing the economic value of non-wood forest ecosystem services by protected areas.
• Section 7: Discussion – summarizes the findings and implications of the report.
• Appendices:
  • Appendix 1: Details the process of updating the metadata analysis.
  • Appendix 2: Lists all studies included and excluded in the meta-analysis.
  • Appendix 3: Describes the study review quality control protocol.
  • Appendix 4: Provides currency conversion information.
  • Appendix 5: Includes statistical methods used in the analysis.
  • Appendix 6: Explains GIS data processing and variables construction.
  • Appendix 7: Contains additional tables and figures.
  • Appendix 8: Lists global predictions by model.
  • Appendix 9: Provides regional and country-level average predictions and total wealth estimates.

Conclusion

The report emphasizes the importance of using local and service-specific valuation methods to avoid bias in global assessments. It provides a detailed framework for understanding the economic value of non-wood forest ecosystem services, highlighting the role of protected areas and the need for improved data representation across different regions and ecosystems.