DASIG model: a data automation program for the calculation of soil trace gas fluxes from paddy rice

We introduce a software application called the DASIG Model, which is a user-friendly tool to automate the calculation of N 2 O, CH 4 , and CO 2 effluxes from rice fields using the chamber method. DASIG's initial release consists of two submodels: FLUX and GWP. The FLUX submodel calculates soil trace gas fluxes from user-provided gas concentrations, determined by traced gas analyzer instruments. DASIG determines gas fluxes from time series of gas concentrations using established trace gas flux models based on linear or non-linear time relationships for paddy rice. It performs additional quality control checks and employs its own decision logic to determine which flux model provides the best fit to sample data. The GWP submodel (Global Warming Potential) converts trace gas flux data into cumulative emissions throughout the crop cycle or crop season, expressed as CO 2 equivalents, using trapezoidal time integration. DASIG was validated using six datasets from various field experimental designs representing management alternatives for paddy rice crops in northeast Arkansas, USA. The model outputs were compared to manual computations of soil trace gas fluxes. Our results showed that soil trace gas flux datasets computed by DASIG model were positively and significantly correlated to manually computed fluxes ( P ≤ 0.0001, R 2 > 0.99). Data processing time was shortened by >1000 times for DASIG model computation compared to manual calculations. DASIG model demonstrates a strict streamlining of trace gas flux calculations that automates and simplifies the quantification of soil trace fluxes, allowing non-experts on biogeochemistry to produce high-quality data. This automated tool is highly relevant in the optimization of uncertainties for accurate inventory of atmospheric trace gases from crop production, validation of soil trace gas models for rice, and as a promising platform to unify trace gas datasets globally. DASIG's modular data processing software is written in Python and is publicly available, distributed as a single installer program that contains all software requirements.