Land surface models (LSM) predict fluxes, and stores of carbon and water in soil and vegetation and are integral to modelling efforts in many domains, e.g. hydrological, climate and all the way up to earth system models. More importantly, quantifying stocks of carbon and water is essential for sustainable management of both natural and agricultural landscapes. The next generation of land surface models must incorporate a new understanding of the function and interaction of soil and plants, but also to incorporate new and rich data streams on land surface change at spatial resolutions relevant to agriculture, the paddock and finer, such as drone and farmer collected yield monitor data.
The generic approach of the land surface modelling community is to focus on big picture, low resolution ecosystem scale fluxes and stores. This results in simplistic assumptions of how ecosystems function which is unsuitable for agricultural applications. This includes simplified landuse classes and continuous management within land use classes. The spatial resolution is typically 5km+, which is well beyond the management resolution for agriculture of a paddock. The next generation of modeling is essential to improve productivity for Australia’s agricultural sector. At the other end of the scale, the agricultural community tends to rely on soil and plant models that require many inputs, and can only be readily applied at few locations where such data is available. Such results are not easily transferrable to the landscape scale where the complexity and data required is not available. Nonetheless, the increasing availability of geospatial data that represents the water and carbon cycle creates the opportunity for a next generation approach through avoiding many of the simplistic assumptions described above