PHOSPHORUS MODELLING

1
PHOSPHORUS MODELLING
Michael Hutchins, Martyn Silgram, Steven Anthony
Eunice Lord Environment Modelling and GIS
Group ADAS Wolverhampton
PHOSPHORUS MODELLING
Environment Modelling and GIS Group
2
INTRODUCTION

• Overall research focus
• 1) The conversion of small scale process
  understanding to large scale decision making -
  focus on aspects of upscaling.
• 2) To develop usable and understandable validated
  models for regional, national and international
  user communities (e.g. implementation of EU WFD).
• Specific uses of models
• Not only for predictive purposes - but also to
  evaluate mitigation strategies.
• Ability to address individual specific issues of
  interest to policy makers, e.g.
• generation of erosion
• within-field transport and export out of field
• impacts of land use change (e.g. winter v.
  spring sown cereals)
• impact of a change in manure/fertiliser
  management practice

3
SITES TO WHICH MODELS APPLIED

• Field / farm scale / small catchment scale
• Titley Court (NT1033), Rosemaund (NT1027)
• Trent (Cliftonthorpe) / Rosemaund (NT1027)
• River basin / National scale
• Swale (NT1024) APECS 2.0
• P Indicators project (PE0105) Windrush,
  Slapton Lea, Bure, Windermere, Waver
• EUROHARP EU 5th Framework study (Yorkshire
  Ouse) APECS 1.0
• Wye and Hampshire Avon (PE0202) - focus on
  mitigation (PSYCHIC)

4
PROCESS ORIENTATED MODELLING APPROACH

• Focus on within-storm dynamics
• Sub-hourly rainfall input

Input data requirements met by national-scale
datasets
EUROSEM
Planar landscape elements and individual fields
Sharpley et al. (1995)

• Modularised modelling involves predictions of
• rainfall intensity-duration profile
• stage of crop development
• soil moisture status
• runoff
• sediment loss
• soil phosphorus mobilisation/desorption
• physical and chemical retention

Hillslope / floodplain and catchment processes
Stream channel
5
CONFIGURATION OF PLANAR ELEMENTS
Stone Farm Grass
5 ha
6 ha
Brushes/Stoney
Longlands
7 ha
4 ha
Foxbridge
Orchard
2 ha
2 ha
2 ha
Stoney
3 ha
Slade Hopyard
Jubilee Station, Rosemaund (30.6 ha silty clay
loam)
6
OBSERVED SEDIMENT AND TP FLUXES (1998-9)
7
EROSION MODEL PERFORMANCE
8
CUMULATIVE LOSSES CLIFTONTHORPE WINTER 98-99
9
CUMULATIVE LOSSES CLIFTONTHORPE WINTER 98-99
10
PROCESS BASED MODELLING CONCLUSIONS

• Successes
• Dynamics of P losses are acute and event based -
  acceptable simulation of erosion-driven
  responses.
• Sensitivity to soil surface conditions (e.g. crop
  growth, seedbed condition, sealing, poaching) can
  be included.
• Further attention needed
• losses via solubilisation
• particulate losses through subsurface pathways
  (at all sites with drains e.g. as observed at
  Titley Court and under grassland at
  Cliftonthorpe)
• the consequences of manure and fertiliser
  additions on soil chemistry and P losses
• retention and temporary retardation both in the
  landscape and in the stream channel.

11
CATCHMENT-SCALE APPROACHES

• Longer time step (e.g. annual predictions),
• Use of grid resolution modelling calculation
  unit (e.g. 1 km2), or field resolution for small
  catchments.
• Risk-based approaches (e.g. indexing schemes)
• Empirical models not requiring prior calibration
  (e.g. P Indicator model, APECS 1.0). Export
  coefficients founded on DEFRA NT10 research
• Make use of national survey/census data (land
  use, agricultural census, soils (including HOST),
  topography, climate - either long term annual,
  year-specific or daily)

12
APECS 1.0 RESULTS YORKSHIRE OUSE
Underestimation at Crakehill and Skelton partly
due to bank erosion, known to be important in
the lower Swale.
13
COMPONENTS TO PHOSPHORUS LOSS

• Areas of high P loss
• urban centres
• areas with high pig and poultry numbers

Variation in HER and soil texture across the
catchment also important and sensitive in
determining total annual loss.
14
FURTHER ISSUES EROSION AT 1 KM2 RESOLUTION

• APECS and P-Indicator models use the Erosion
  Risk Classification (NSRI)
• a categorisation (based on soil texture, crop
  type, annual rainfall)

• P export coefficients
• are assigned for each
• of the 6 classes.
• The coefficients are
• modified by a crop risk
• factor

• Alternative generate a continuum of erosion
  loss, include all land uses.
• Quantitative, multiplicative treatment of
  factors representing key components (e.g. soil
  erodibility, effect of crop cover).
• Or, make use of generalisation of EUROSEM
  (response surfaces).
• Combine with description of desorption kinetics
  of P (PSYCHIC).

15
EROSION RESPONSE SURFACE

• Surface specific for
• soil texture
• soil moisture status
• slope length and angle
• rainfall input

• Summarises effects of
• cropping
• management
• seasonality

May
October
EUROSEM RFR parameter (surface roughness)
Fractional crop cover
16
TOWARDS UNDERSTANDING FACTORS CONTROLLING
TRANSFER OF PHOSPHORUS WITHIN AND FROM
AGRICULTURAL FIELDS
Mike Hutchins, Robin Hodgkinson, Bob Jackson
Eunice Lord (ADAS) John Quinton (NSRI, Silsoe)
DEFRA project PE0111 18 months (ends June
2003) 125k
17
SCALING UP FROM PLOT TO WHOLE-FIELD

• CSG7 Objectives
• Characterise water, sediment and P loss response
  across a range of slope (angle, length and form)
  attributes.
• Derive rules expressing these effects for use in
  catchment-scale modelling, by developing/applying
  standardised scenarios using existing
  process-orientated models.
• Test models with data collected to reconcile
  measurements and derived rules. Hence a validated
  general predictor of delivery within field can be
  developed, suitable for use in catchment scale
  models (e.g. PE0202).
• Are integrated, whole-field losses equal to the
  sum of losses from planar components?

18
SCALING UP FROM PLOT TO WHOLE-FIELD

• Techniques
• Field runoff trap approaches (ADAS) at
  Rosemaund Longlands
• runoff traps 3 slope lengths (10-100m)
  with/without tramlines
• edge of field gutter
• Triaxial laboratory soil flume (NSRI, Silsoe)
• 3 slope forms (planar, concave and convex)
• Additional method development spatially
  extensive sampling (Hattons)
• low cost
• unobtrusive to farming operations
• large number of integrated samples
• Modelling
• field scale EUROSEM (erosion) and Sharpley
  algorithms (P dynamics)
• results inform catchment scale (P Indicator,
  PSYCHIC)