Introduction
The EFAS and GloFAS soil water content is modelled in LISFLOOD in three layers representing top, medium and bottom sections of the soil (indicated as superficial, upper and lower soil in Figure 1). Further, for each LISFLOOD pixel, the soil moisture calculation is linked to land use classes by using the main fractions of forest (f), irrigation (i) and other (o). The fractions of each land use type do not necessarily add up to 1 for each pixel since they represent the soil area that is not water for each grid cell.
The LISFLOOD output soil moisture variable is volumetric soil water content, with a separate variable existing for the forest, irrigation and other land types, for the three layers of the soil separately for all three fractions. All these are defined as the ratio between the volume of water and the total volume of the soil, including all particles, vegetation, water and air.
The soil wetness index (swi) is defined as a summary variable, representing the soil moisture conditions in the so-called root zone. The root zone is the maximum depth at which plants can extract water from the soil, and as such helps with interpretation of the soil moisture, especially across different models with different discretisation of the soil. In LISFLOOD, the root zone is represented by the top two soil layers. The 1st layer has a fixed 50 mm thickness, while the 2nd layer has variable thickness across the different land fractions.
The soil wetness index is a value from 0 to 1, computed by scaling the actual root zone soil water content (a ratio between the actual soil water content volume in the root zone over the total volume of the soil) between the residual soil water content and saturation soil water content, as the lower and upper extremes of the root zone.
Figure 1. Representation of the hydrological processes in LISFLOOD with the three soil layers (from OS LISFLOOD manual at https://ec-jrc.github.io/lisflood-model/).
In the soil wetness index definition used here, the 'irrigation' fraction of the soil moisture is not directly considered and only represented as part of the 'other' soil fraction. This is an acceptable simplification as the irrigation fraction has only small contribution in general.
Mean root zone soil water content for actual, residual and saturation computation
Step one
The actual (swact), the residual (swres) and saturation (swsat) soil water contents are computed over the whole root zone, as a weighted average over the two soil fractions and two soil layers:
- swact = (ff*(swaf1*sdf1+swaf2*sdf2) + ff1*(swao1*sdo1+swao2*sdo2)) / (ff*(sdf1+sdf2)+ff1*(sdo1+sdo2))
- swres = (ff*(swrf1*sdf1+swrf2*sdf2) + ff1*(swro1*sdo1+swro2*sdo2)) / (ff*(sdf1+sdf2)+ff1*(sdo1+sdo2))
- swsat = (ff*(swsf1*sdf1+swsf2*sdf2) + ff1*(swso1*sdo1+swso2*sdo2)) / (ff*(sdf1+sdf2)+ff1*(sdo1+sdo2))
Step two
The actual root zone soil water content is scaled linearly between the residual and saturated root zone water contents:
- swi = (swact-swres) / (swsat-swres)
Availability of the soil moisture and related data
Soil moisture information is available though MARS and CDS. The soil wetness index (root zone) is provided from EFAS v5 and GloFAS v4, while the three individual soil layer water contents were provided only for EFAS v2, v3 and v4.