Soil respiration pulses in response to pulsed wetting ("Birch effect"; Birch 1958) have long been observed from laboratory and field experiments. The Birch effect produces more CO2 efflux and sustains greater microbial biomass than constantly moist soils. Various mechanisms causing the effect have been proposed. However, the exact mechanism underlying the Birch effect is not clear, and thus most models are not able to simulate this effect. We have recently developed a microbial enzyme based decomposition and gas transport model. The model integrates the most recent advances in the understanding of critical processes, including enzyme-catalyzed degradation of soil organic carbon (SOC) to dissolved organic carbon (DOC), acclimation of carbon use efficiency (CUE) for the uptake of DOC by microbes, and diffusive and convective transport of O2 and CO2 in the soil. The model has four kinds of carbon pools including SOC, DOC, microbial biomass (MIC), and extracellular enzyme (ENZ). However, the model coupled with a land surface model, which accurately simulates soil moisture and temperature, failed to simulate the Birch effect observed at a natural savannah ecosystem site in the southwest US monsoon region. We further divided the DOC and ENZ pools into two sub-pools, one for a wet zone and the other for a dry zone, respectively. We assume that in the dry zone, DOC can be produced through enzyme catalysis, although at a lower rate due to enzyme immobilization, and only in the wet zone can microbes take up DOC. Thus, the modeled DOC accumulates during dry periods and is quickly transitioned into DOC in the wet zone (proportional to saturation) in response to pulsed wetting during a storm, and becomes available for microbial use. In such a way, the model successfully simulates the Birch effect with the Nash–Sutcliffe model efficiency being ~ 0.75 (correlation coefficient ~ 0.88) at a half-hourly time step. We will also present the effect of gas transport on the Birch effect and the Birch effect on long-term soil carbon dynamics.