me3205 – ecosystem connections...

7
1 ME3205 – Ecosystem Connections project Report - Impact of climate variability and trawling on ecosystem dynamics: GIS and ERSEM model analysis Appendix 2: ERSEM 3D output model validation – temporal variance, site fluxes of nutrients and nutrient profiles 1 Validation of 3D ERSEM 3D ERSEM outputs of seasonal nutrient fluxes at the benthic-pelagic interface (ambient) and nutrient content of the bed for the period 1997-2003 were compared with measurements from various NERC and Defra projects, and in relation to the main MEC sites. The purpose was to decide whether the model outputs could be used to provide the nutrient fluxes and nutrient content of the bed which would be used as baseline or ambient within the natural and trawling disturbance comparisons. Validation of the 3D outputs focused on nutrient fluxes at the benthic-pelagic interface (inter- annual variability, annual average and seasonality), pore-water concentrations and profiles. A detailed comparison is made below. Overall, the modelled nutrient flux and sediment profile data compared well enough with the observations to use the model output for the calculation of annual total nutrient fluxes and to use the regional differences for the natural and trawling disturbance calculations. The largest discrepancies between the model results and the observations occurred for the Sean Gas Field in the Southern Bight and site north of Dogger Bank, where concentrations of nutrients within the bed were over and underestimated respectively. At the Sean Gas Field in the Southern Bight the over-estimation is largely due to the model not seeing this area as an advective type substrate which acts to equilibrate the pore-water nutrient concentration in the upper layers of the sediment with the water column. At the site north of Dogger Bank, the under-estimation is a result of a smaller modelled organic matter (and hence nutrient) supply to the bed. Overall, the seasonal average pore-water concentrations are reasonably well predicted by the model at all sites, with the exception of overestimation of NH4 and PO4 concentrations at the Oyster Grounds. All these issues are discussed in more detail in the 1D model validation report (Coughlan et al., 2010). The results are interpreted with these compromises in mind. The annual cycles for the period 1997-2003 of the fluxes at the benthic-pelagic interface of the four nutrients in the model show considerable interannual variability, in particular for the summer (Figure 1). Because the field observations were taken in 2007 and 2008, for which model results were not available, 7-year averages and variances were calculated from the modelled fluxes and compared with the observations. The averaged modelled nutrient fluxes were generally within the variance of the observations (Figure 2 and 3), with the exception of phosphate and ammonium fluxes for the Oyster Grounds (over-estimated by the model). Comparison of the vertical distributions of pore-water nutrients in the model with observations (Figure 4) showed that concentrations at the Sean Gas Field in the southern Bight were generally over-estimated due to the lack of physical advection in the bed. At the site north of Dogger Bank and the Oyster Grounds, the shape of the profiles was represented well by the model, but concentrations were mostly over-predicted. Also, the model yielded larger concentrations at the Oyster Grounds than at the site north of Dogger Bank, whereas the observations indicated larger concentrations at the site north of Dogger Bank.

Upload: others

Post on 08-Jul-2021

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

1

ME3205 – Ecosystem Connections project Report - Impact of climate variability and trawling on ecosystem dynamics: GIS and ERSEM model analysis

Appendix 2: ERSEM 3D output model validation – temporal variance, site fluxes of nutrients and nutrient profiles

1 Validation of 3D ERSEM 3D ERSEM outputs of seasonal nutrient fluxes at the benthic-pelagic interface (ambient) and nutrient content of the bed for the period 1997-2003 were compared with measurements from various NERC and Defra projects, and in relation to the main MEC sites. The purpose was to decide whether the model outputs could be used to provide the nutrient fluxes and nutrient content of the bed which would be used as baseline or ambient within the natural and trawling disturbance comparisons. Validation of the 3D outputs focused on nutrient fluxes at the benthic-pelagic interface (inter-annual variability, annual average and seasonality), pore-water concentrations and profiles. A detailed comparison is made below. Overall, the modelled nutrient flux and sediment profile data compared well enough with the observations to use the model output for the calculation of annual total nutrient fluxes and to use the regional differences for the natural and trawling disturbance calculations. The largest discrepancies between the model results and the observations occurred for the Sean Gas Field in the Southern Bight and site north of Dogger Bank, where concentrations of nutrients within the bed were over and underestimated respectively. At the Sean Gas Field in the Southern Bight the over-estimation is largely due to the model not seeing this area as an advective type substrate which acts to equilibrate the pore-water nutrient concentration in the upper layers of the sediment with the water column. At the site north of Dogger Bank, the under-estimation is a result of a smaller modelled organic matter (and hence nutrient) supply to the bed. Overall, the seasonal average pore-water concentrations are reasonably well predicted by the model at all sites, with the exception of overestimation of NH4 and PO4 concentrations at the Oyster Grounds. All these issues are discussed in more detail in the 1D model validation report (Coughlan et al., 2010). The results are interpreted with these compromises in mind. The annual cycles for the period 1997-2003 of the fluxes at the benthic-pelagic interface of the four nutrients in the model show considerable interannual variability, in particular for the summer (Figure 1). Because the field observations were taken in 2007 and 2008, for which model results were not available, 7-year averages and variances were calculated from the modelled fluxes and compared with the observations. The averaged modelled nutrient fluxes were generally within the variance of the observations (Figure 2 and 3), with the exception of phosphate and ammonium fluxes for the Oyster Grounds (over-estimated by the model). Comparison of the vertical distributions of pore-water nutrients in the model with observations (Figure 4) showed that concentrations at the Sean Gas Field in the southern Bight were generally over-estimated due to the lack of physical advection in the bed. At the site north of Dogger Bank and the Oyster Grounds, the shape of the profiles was represented well by the model, but concentrations were mostly over-predicted. Also, the model yielded larger concentrations at the Oyster Grounds than at the site north of Dogger Bank, whereas the observations indicated larger concentrations at the site north of Dogger Bank.

Page 2: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

2

Figure 1: 3D ERSEM output annual time series fluxes between pelagic and benthic systems at the three sites (mmol/m2/d), monthly averages 1997-2003 for N1p (PO4), N3n (NO3), N4n (NH4) and N5s (dissolved Si).

Page 3: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

3

Page 4: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

4

Figure 2: Seven-year average model flux at the Oyster Grounds (OG), site north of Dogger Bank (ND) and the Sean Gas Field in the Southern Bight (SG) (1997-2003 in blue) for P, NO3, NH4 and dissolved Si – MECS obs data are displayed in green, mean of all observational data (yellow) and mean of observational data for corresponding silt range (OG, ND or SG in red)+/- standard deviation.

Page 5: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

5

Page 6: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

6

Figure 3: Overall annual average of model and MECS observational data +/- standard deviation

Page 7: ME3205 – Ecosystem Connections projectrandd.defra.gov.uk/Document.aspx?Document=ME3205Appendix... · 1 ME3205 – Ecosystem Connections project Report - Impact of climate variability

7

Figure 4: Sediment profiles of nutrients and MEC observational comparisons: Seabed nutrient profiles (modelled and observed for main hydrodynamic/seabed regions) Figure 1: Model output of nutrient profiles within the main hydro/seabed regions (SB, OG and ND) – mmol/m-3 pw.

Figure 5: Observational pore-water nutrient (µmol/l) profiles from three main regions for a) May 2007 and b) seasonal averages