Key points are not available for this paper at this time.
Rice cultivation is the largest source of CH₄ emissions in Vietnam's agricultural sector, making robust quantification essential for national GHG inventories and mitigation planning. This study compares three IPCC approaches for estimating CH₄ emissions from Vietnamese rice systems: Tier 1 (global emission factors, EFs), Tier 2 (national EFs), and Tier 3 (two process-based biogeochemical models, DNDC and LandscapeDNDC). Model estimates were evaluated against a site-level CH₄ dataset, grouped by the availability of local soil data and full-year management records. Tier 1 underperformed across all sites. Tier 2 showed moderate and relatively stable agreement with observations for both the full dataset (R² = 0.31; NSE = 0.27) and better-documented sites (R² = 0.27; NSE = 0.21). In contrast, Tier 3 accuracy depended strongly on information from both the evaluated and preceding season. For DNDC, NSE (R²) improved from 0.25 (0.30) at data-limited sites to 0.41 (0.47) at better-documented sites; for LandscapeDNDC, NSE (R²) increased from 0.24 (0.44) to 0.65 (0.71). Under complex water management (e.g. multiple drainage events), Tier 3 performance declined; a simple input perturbation showed that uncertain soil hydrologic properties strongly affect simulated CH₄, making it difficult to determine whether the reduced performance reflects model limitations or input uncertainty. For inventory design, Tier 2 currently offers a practical balance between accuracy and data requirements and should remain the backbone of national reporting. Targeted Tier 3 applications can add value where full-year management information and basic soil data are available, especially for analysing temporal emission dynamics and evaluating mitigation options.
Nguyen et al. (Fri,) studied this question.