National estimates of phosphorus (P) losses from Swedish agricultural land to surrounding
waters are currently produced using the calculation tool ICECREAMDB
and its constituent simulation model ICECREAM. The ICECREAM model has its
origins in the American models CREAMS and EPIC, and parts of it have been
adapted for calculating the solubility of P under Finnish conditions. It would be
preferable if the model were better adapted for the soil conditions and soil mapping
data available in Sweden. In the present work we focused on the model’s description
of inorganic soil P, its storage in different pools and its solubility. We identified
the problem that the model in its current form is slightly too rigid and insensitive
to changes in fluctuations in more mobile P fractions. The model is unable to
reproduce either the within-year dynamics or the long-term dynamics of a more
easily soluble, plant-available pool, since that pool and its associated flows are too
small in relation to the large total-P pool. Since the parameterisation of pools and
flows of inorganic P is not adapted to the soil analysis methods used in Sweden, the
availability of suitable indata is limited.
Recent research in Sweden has revealed a correlation between the P content of the
soil extracted with ammonium lactate and expressed as P-AL and soluble P in the
topsoil extracted with CaCl2 or H2O. The results are not as clear regarding the P
status of the soil and soluble P in drainage water or in neighbouring water courses.
This is because other characteristics, both chemical and physical, and how these are
distributed throughout the entire soil profile must be taken into account. The research
has also shown that taking into account the P sorption capacity of the soil
and the degree of P saturation can improve the correlation between soil P status and
dissolved P for certain soils. If these correlations can be verified for a larger number
of soils, it may be worthwhile including them in the simulation model in order
to improve calculations of P losses.
There is a need for improved models that reflect the latest research and that modify
the area of application of the models from the original intention. Such models will
need to include additional or modified descriptions of various processes. It would
be desirable to base the description of the soil’s inorganic P pools in ICECREAM
on the analytical methods used routinely in Sweden. It would also be desirable to
take into account additional chemical characteristics of the soil in calculating the
balance and thereby the flows between the different P pools in the model. Future
research will show whether this is possible. This report presents a review of existing
knowledge and should be regarded as decision support for future planning for
such developments. The research results contained in the report must be verified
for a greater number of soils and model development must be tailored to match the
available indata. The underlying motive for the development must be a need to
describe the effects of cultivation measures that affect P dynamics, which we are
unable to do at present.
Norrköping: SMHI , 2011. , p. 36