Bathtub Model Equations &
Options
Conservative Substance Balance | |
Option | Description |
0 | Do Not Compute (Set Predicted = Observed) [default] |
1 | Compute Mass Balances |
Phosphorus Sedimentation Models (see discussion) | |||
Unit P Net Sedimentation Rate (mg/m3-year) = CP A1 PA2 Solution for Mixed Segment: Second-Order Models (A2 = 2): P = [-1 + (1 + 4 K A1 Pi T)0.5 ] / (2 K A1 T) First-Order Models (A2 = 1):
P = Pi / (1 +
K A1 T) |
|||
Option | Model Description | A1 | A2 |
0 | Do Not Compute (Set Predicted = Observed) [default] | - | - |
1 | Second-Order, Available
P
[default] Inflow Avail P = 0.33 Pi + 1.93 Pio See options for specification of available P |
0.17 Qs/(Qs + 13.3) |
2 |
2 | Second-Order Decay
Rate Function Fot = Tributary Ortho P / Total P Load Requires specification of inflow total & ortho P loads |
0.056 Fot-1Qs/(Qs
+ 13.3) Qs = Max(Z/T,4) |
2 |
3 | Second-Order | 0.10 | 2 |
4 | Canfield & Bachman (1981), Reservoirs |
0.114 (Wp/V)0.589 |
1 |
5* | Vollenweider (1976), Northern Lakes |
T-0.5 |
1 |
6* | Simple First-Order |
1 |
1 |
7* | First-Order Settling | 1/Z | 1 |
8* | Canfield & Bachman (1981), Natural Lakes |
0.162 (Wp/V)0.458 |
|
9* | Canfield & Bachman (1981), Reservoirs + Lakes |
0.129 (Wp/V)0.549 |
1 |
For purposes of computing effective rate coefficients (A1), Qs, Wp, Fot, T, and V are evaluated separately for each segment group based upon external loadings and segment hydraulics.
* These models are not calibrated to CE reservoir data. They are likely to require calibration by the user to site-specific data.
Nitrogen Sedimentation Models (see discussion) | |||
Unit N Net Sedimentation Rate (mg/m3-year) = CN B1 NB2 Solution for Mixed Segment: Second-Order Models (B2 = 2): N = [-1 + ( 1 + 4 K B1 Ni T )0.5 ] / ( 2 K B1 T ) First-Order Models (B2 = 1):
N = Ni / (1 + K B1 T) |
|||
Option | Model Description | B1 | B2 |
0 | Do Not Compute (Set Predicted = Observed) [default] | - | - |
1 | Second-Order, Available N
[default] Inflow Avail N = 0.59 Ni + 0.79 Nin See options for specification of available N |
0.0045 Qs/(Qs +
7.2) |
2 |
2 | Second-Order Decay
Rate Function Fin = Tributary Inorganic N/Total N Load Requires specification of inflow total & inorganic N loads |
0.0035 Fin-1Qs/(Qs
+ 17.3) Qs = Max(Z/T,4) |
2 |
3 | Second-Order |
0.00315 |
2 |
4* | Bachman (1980), Volumetric Load |
0.0159 (Wn/V)0.59 |
1 |
5* | Bachman (1980), Flushing Rate |
0.693 T-0.55 |
1 |
6* | Simple First-Order |
1 |
1 |
7* | First-Order Settling | 1/Z | 1 |
Nitrogen Model 1 differs slightly from that developed in Walker (1985). The coefficients have been adjusted so that predictions will be unbiased if inflow inorganic nitrogen data are not available (inflow available N = inflow total N). These adjustments have negligible influence on model error statistics< /A > .
For purposes of computing effective rate coefficients (B1), Qs, Wn, Fin, T, and V are evaluated separately for each segment group based upon external loadings and segment hydraulics.
* These models are not calibrated to CE reservoir data. They are likely to require calibration by the user to site-specific data.
Application of Nutrient Availability Factors (see discussion) | ||
Option | Description | Equations |
0 | Do Not Apply Availability Factors [default] |
Select if Ortho P & Inorganic N loadings are not specified Inflow Available P = Pi Inflow Available N = Ni |
1 | Apply to P & N Sedimentation Model 1 Only |
When P Model 1 or N
Model 1 is selected, calculate nutrient balances based upon available
nutrient loads:
Inflow Available P =
0.22 Pi + 1.93 Pio Require specification of ortho P &
inorganic N loadings if P or N sedimentation model 1 is selected.
|
2 | Apply to All P & N
Sedimentation Models Except Model 2 |
Same equations as Option 1 Requires Specification of Ortho P & Inorganic N Loadings |
Account for differences between dissolved &
particulate nutrient forms with respect to sedimentation rates and/or
bioavailability.
Although consideration of these factors reduced prediction
error in the CE model development data set, the default
option (0) ignores them because
ortho P and inorganic N loadings are not typically measured.
Applicability Constraints | ||||||
Option | Description / Limiting Factors | Equations | a | (N-150)/P | Ninorg/Portho | Fs |
0 | Do Not Compute | Predicted = Observed | ||||
1 | P, N, Light, Flushing |
Xpn = [ P-2
+ ((N-150)/12)-2 ]-0.5 |
||||
2 | P, Light, Flushing [default] |
Bp =
P1.37/4.88
|
>12 | >7 | ||
3 | P, N, Low Turbidity | B = K 0.2 Xpn1.25 | <0.9 | <25 | ||
4 | P, Linear | B = K 0.28 P | <0.9 | >12 | >7 | <25 |
5 | P, Exponential, Jones & Bachman (1976) | B = K 0.081 P1.46 | <.4 | >12 | >7 | <25 |
6 | P, Carlson TSI (1977), Lakes | B = K 0.087 P1.45 | <0.4 | >12 | >7 | <25 |
Options 1 & 2 require estimates of non-algal turbidity for each model segment. These are entered with observed water quality data on the 'Edit Segments' screen. If non-algal turbidity is not specified, it is estimated from observed Secchi depth and chlorophyll-a. If the latter are not specified, an error message is generated.
Secchi Depth Models (See discussion) | Applicability Constraints | |||
Option | Description | Equations | (N-150)/P | Ninorg/Portho |
0 | Do Not Compute | Predicted = Observed | ||
1 | Secchi vs. Chl a and Turbidity [default] | S = K / (a + b B) | ||
2 | Secchi vs. Composite Nutrient | S = K 16.2 Xpn-0.79 | ||
3 | Secchi vs. Total P, CE Reservoirs | S = K 17.8 P -0.76 | >12 | >7 |
4* | Carlson TSI (1977) , Lakes | S = K 48 / P | >12 | >7 |
Longitudinal Dispersion Models (see discussion) | ||
Option | Description | Equations |
0 | Do Not Compute | E = 0 |
1 | Fischer et al.
(1979) Dispersion Equation as adapted by Walker (1985) [default] |
Width W = As/L Cross-Section Ac = W Z Velocity U = Q/Ac Dispersion D = KD 100 W2 Z-0.84 Maximum (U,1) Numeric Dispersion Dn = U L/2 Exchange E = MAX( D - Dn , 0 ) Ac/L |
2 | Fixed Dispersion Rate |
Same as Model 1,
except with fixed D = 1,000 km2/year D = 1000 KD |
3 | Input Exchange Rates Directly | E = KD |
4 | Fischer Equation, Not Adjusted for Numeric Dispersion | E = D Ac/L (D as defined in Option 1) |
5 | Constant Dispersion, Not Adjusted for Numeric Dispersion | E = 1,000 KD Ac / L |
Estimate Exchange Flows (E) between Adjacent
Segment Pairs.
For all options, E = 0. always for segments discharging out of network
(outflow segment number = 0).
Error Analysis (see discussion) | |
Option | Description |
0 | Do Not Perform (Output CV's = 0) |
1 | Consider Model Error & Input Error [default] |
2 | Consider Model Error Only (reflect inherent model error only) |
3 | Consider Input Error Only (reflect uncertainty in user-specified inputs only) |
Application of Phosphorus Calibration Factors (see discussion) | |
Option | Description |
0 | Apply Calibration Factors to Predicted Sedimentation Rates [default] |
1 | Apply Calibration Factors to Predicted Concentrations |
Application of Nitrogen Calibration Factors (see discussion) | |
Option | Description |
0 | Apply Calibration Factors to Predicted Sedimentation Rates [default] |
1 | Apply Calibration Factors to Predicted Concentrations |
Calculation of Mass Balance Tables (see discussion) | |
Option | Description |
0 |
Use Observed Segment
Concentrations to Calculate Outflow and Storage Terms Mass balance tables are based entirely on observed inflow, outflow, & segment concentrations specified by the user (independent of predicted reservoir or outflow concentrations). If observed outflow or segment concentrations are missing, predicted values are used. |
1 | Use Predicted Segment Concentrations to Calculate Outflow and Storage Terms [default] |
Influences output from the 'List Balances ' procedures. Does not influence predicted concentrations.
Equations for Other Trophic Response & Morphometric Variables (see discussion & output variable descriptions) | |
Variable |
Equations |
Organic Nitrogen |
Norg = K ( 157 + 22.8 B + 75.3 a ) |
Total P - Ortho P |
P - Portho = K Maximum [ -4.1 + 1.78 B + 23.7 a , 1 ] |
Hypolimnetic Oxygen Depletion Rate (Near-Dam) |
HODv = K 240 Bm.5 / Zh (for Zh > 2 m) |
Metalimnetic Oxygen Depletion Rate (Near-Dam) |
MODv = K 0.4 HODv Zh 0.38 |
Principal
Components
With Chl-a, Secchi, Nutrient, & Organic N Data |
PC-1 = 0.554 log(B) +
0.359 log(Norg) + 0.583 log(Xpn) - 0.474 log(S) PC-2 = 0.689 log(B) + 0.162 log(Norg) - 0.205 log(Xpn) + 0.676 log (S) |
Principal
Components With Chl a & Secchi Data Only |
PC-1 = 1.47 + 0.949
log(B) - 0.932 log(S) PC-2 = 0.13 + 0.673 log(B) + 0.779 log(S) |
Trophic State Indices (Carlson 1977) |
TSIp = 4.15 + 14.42 ln(P)
TSIc = 30.6 + 9.84 ln(B) TSIs = 60.0 - 14.41 ln(S) |
Algal Bloom Frequencies
(Walker 1984) |
Calculated from Mean
Chl-a (B) assuming that temporal variations in chl-a |
Non-Algal Turbidity |
a = 1/S - b B
(minimum value = 0.08 1/m) Applied to observed S and B values in each segment if non-algal turbidity values are not input directly (=0) on the Edit Segments screen. The parameter b (default = 0.025) is entered on the Model Coefficients screen. |
Mean Depth of Mixed Layer |
log (Zmix) =
-0.06 + 1.36 log (Z) - 0.47 [log (Z)]2 (R2 =
0.93, SE2 = 0.0026) |