Large-scale prediction of organic matter decomposition potential using machine learning and field decomposition data of tea-bag and chotsticks

2022 Fiscal Year Final Research Report

• Project/Area Number: 19K15879
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 40010:Forest science-related
• Research Institution: Forest Research and Management Organization
• Principal Investigator: Mori Taiki 国立研究開発法人森林研究・整備機構, 森林総合研究所, 主任研究員 等 (90749095)
• Project Period (FY): 2019-04-01 – 2023-03-31
• Keywords: 有機物分解 / 気候変動 / 機械学習 / ティーバッグ法

Outline of Final Research Achievements

A prediction model for tea bag decomposition rate was established using the random forest method. Additionally, a modification was proposed to improve the prediction accuracy of the tea bag index (TBI) approach, which is a standard method for assessing litter decomposition. These contributions help to improve the robustness of the impact assessment of soil organic matter decomposition under climate change. Furthermore, we accumulated basic data on the effect of soil water on the decomposition rate of tea bags, revealing that the effect of soil moisture on tea bag decomposition rate is nonlinear. We also demonstrated through field fertilization experiments that carbon addition accelerates tea bag decomposition more than nutrient addition.

Free Research Field

生態学

Academic Significance and Societal Importance of the Research Achievements

ランダムフォレスト法によるティーバッグ分解速度予測モデルを作成したところ、平均気温、降水量、日射量、相対湿度が上昇するほど、また、標高、傾斜、Topographic Position Indexが低下するほどティーバッグの分解が速くなるという従来の知見と調和的な結果が得られ、機械学習によるリター分解速度予測マップ作成が十分可能であることが明らかになった。これは炭素排出量の算定方法改善に貢献する。