Variations in gas transport parameters at the field scale govern the transport, fate and emission of greenhouse gases and volatile organic chemicals in soil. In this study, we analyzed spatial variability and evaluated predictive models for soil-gas diffusivity (D_p) and air permeability (k_a) based on measurements along a 117-m long transect and a parallel 33-m short transect of a volcanic ash soil (Andisol) in Nishi-Tokyo, Japan. Measurements were done on 100-cm³ undisturbed soil samples, with 3-m spacing between sampling points, and included water retention, D_p and k_a at different soil-water matric potentials, and saturated hydraulic conductivity (K_s). Traditionally-used gas diffusivity models including the Millington-Quirk and Buckingham-Burdine-Campbell (BBC) models underestimated D_p at field capacity moisture content (~-100 cm H₂O matric potential) and largely over-estimated D_p under very dry conditions, likely due to effects of soil aggregation. A simple linear model (Penman-Call) for D_p as a function of air-filled porosity (e), taking into account inactive air-filled pore volume (e_in), accurately described D_p from wet to oven-dry conditions, and well captured the spatial variations in D_p along the transects. Air permeability also exhibited an almost linear increase with e but data showed no evidence of e_in effects, and k_a was best predicted from a newly presented power-law model (Kawamoto-Alexander) with k_a at -100 cm H₂O matric potential as the measured reference point. Trends of decreasing soil-water retention and increasing e and D_p were observed along the transects. Similar trends in k_a and K_s were not observed, likely due to that the convective fluid transport parameters were mainly governed by soil structure instead of fluid phase contents. Autocorrelograms suggested a spatial correlation range of 10-20 m for all three transport parameters (D_p, k_a, K_s). Also, K_s was highly correlated with k_a at -100 cm H₂O matric potential. Measurements of e and k_a at around -100 cm H₂O matric potential (field capacity moisture conditions) are recommended for rapid assessment of magnitude and spatial variations in gas and water transport properties at the field scale.