flowchart LR PPDB[(Hertfordshire<br>Pesticide Properties<br>Database)] --> PersistMobiity[Half life<br>sorption KOC] Labels[(Pesticide<br>labels)] --> AppRate[Application<br>rate] Interview[(Cooperator<br>interview)] --> AppRate SoilSurvey[(NRCS Soil<br>Survey)] --> Soil[Fraction organic carbon<br>bulk density] OurSoilSamples --> Soil Lab[(DEC<br>Air<br>lab)] --> DetLim[Analytical<br>detection<br>limit] DB[(Project<br>database)] --> Times[Sampling<br>times] Assumptions[Assumed<br> - Dist zone thickness<br> - Frac preferential] DB --> Detections[Pesticide<br>detections<br>in samples] TGUS((Compute<br>TGUS<br>index)) PersistMobiity --> TGUS AppRate --> TGUS Soil --> TGUS DetLim --> TGUS Times --> TGUS Assumptions --> TGUS TGUS --> Compare((compare index<br>to detection status)) Detections --> Compare
HowTo – TGUS leaching model (30%)
Priority: low (modeling is low priority)
Updating: rare
This Howto covers one way in which the project compares detections of different pesticides and metabolites at different categorical sites, taking account of different soil, pesticide fate, and management characteristics.
Change log:
| When | Who | Comment |
|---|---|---|
| 2024 05 03 | Sp17 | Began. |
| 2024 07 18 | Sp17 | Fleshing out. Added an appendix with equations and variables. |
| 2024 08 27 | Sp17 | Paper has been published and we are working on the annual report drawing from paper. Working toward a somewhat simplified version of the material to reach a broader audience. |
Related howtos:
1. Background
1.1 Physical causality of pesticides reaching groundwater monitoring points
For a pesticide applied at the land surface to reach groundwater, it must both survive degradation and move downward, and sometimes move sideways in the saturated zone to reach a monitoring point. For a concentration to be observable in groundwater it must also be above an analytical detection limit.
The “Theoretical GUS” (TGUS)1 method, inspired by Gustafson’s Groundwater Ubiquity Score (GUS)2 method for estimating the relative likelihoods of different pesticide residue occurrence in groundwater, computes index scores for different pesticides in a specified soil and management environment such as a cooperating categorical site for the project, and with a specified analytical detection limit. The scores are sensitive to eight parameters, six of which may be quantified. Where GUS is only about a chemical, TGUS is about a chemical, its management, a site, and the analytical detection process.
TGUS applies to a pesticide that is spread mostly uniformly over an area such as a field. It does not apply to concentrated leaks and spills, or to spot applications of pesticides such as injections into trees. It does apply to a pesticide applied to several lines of an area (such as vine rows or potato furrows) rather than the entire area, though this may violate some conceptual model assumptions.
The TGUS index score has units, in days, and the value means how long an applied pesticide would remain above the analytical detection limit after it has been leached out of the near-surface by a mobilizing rain or irrigation event.
1.2 Conceptual leaching model
We begin with a preferental pathway leaching model that consists of two soil strata, a thin upper stratum to which the pesticide is applied and through which it spreads, called a distribution zone. Beneath the distribution zone and extending to the water table is a transmission zone. The soil and pesticide interact intimately within the distribution zone, including sorbing to and desorbing from the soil matric organic matter. The soil and pesticide do not interact with the vadose zone matrix, instead passing through it very quickly in response to a leaching event that fills preferential flow paths.
In addition, the vadose zone’s matrix will carry recharging water but no pesticides. The water via the matrix dilutes the concentration at the water table.
This conceptual model is for one pesticide application, only if there is a hydrological leaching event shortly after the pesticide is applied. The event must be of a size to fill large soil voids near the surface with water and mobilize solutes downward. Intense rain events are increasingly common in Upstate New York (e.g. https://ny-idf-projections.nrcc.cornell.edu/ ).
1.3 Parameters of conceptual model
Pesticide detections near to a pesticide use area, after a leaching event, are a function of several aspects of the pesticide, how it was used, the site, and an analytical detection context:
relative application rate and detectability parameters:
pesticide application rate per unit area
analytical detection limit
thickness of distribution zone (assumption)
chemical fate parameters:
soil half life
organic carbon partition coefficient (KOC)
soil parameters (distribution zone):
bulk density
organic carbon fraction
transmission (vadose) zone parameters:
- fraction of area that consists of connected preferential pathways (assumption)
Notably absent are the depth to groundwater, though the conceptual model’s assumption of very rapid transport implies that the vadose zone is thin enough to have continuous preferential flow pathways to the water table.
For the Upstate NYSDEC monitoring project, we can estimate or measure all but the two assumed values. For certain categories of categorical sites who use pesticides outdoors, on relatively uniform areas as opposed to point applications, we combine cooperator interviews, pesticide labels, soil survey information, our own soil tests, lab-specified detection limits, and literature values.
(NOTE: needs work)
2. Objectives
- Organize systematically the parameter values for sites, pesticides used, and detection limits.
- Attempt to explain causal factors behind pesticide and metabolite detections at categorical groundwater sites.
- Consider timings of semi-annual monitoring.
3. Input data
(how to estimate each parameter)
(sensitivity analysis ranges)
4. Software
(probably a spreadsheet)
Appendix: Equations from the TGUS model
These are from a paper by Steenhuis, project staff and a group alumna 3. See the paper for units and dimensions of variables and parameters.
| Equation, variables, parameters | Equation number in journal article |
|---|---|
| GUS = log(t_{1/2})(4 - log(K_{oc})) | (eq 1a) |
GUS : Gustafson’s groundwater ubiquity score K_{oc} : Organic carbon partition coefficient for the pesticide, representing the partition between dissolved and sorbed t_{1/2} : Pesticide degradation half life in soil |
|
| GUS = log(t_{1/2}) log(\frac{10^4}{K_{oc}}) | (eq 1b) |
| C^w_{dist} = C^w_0 exp(-\frac{q t}{W_{dist}}) | (eq 2) |
C^w_{dist} : concentration of mobile pesticide in water of distribution zone, at time t C^w_0 : initial total concentration of mobile pesticide in water of distribution zone, at time of application q : recharge flow rate in distribution zone, depth of water per unit time t : time elapsed since leaching event after pesticide application |
|
| W_{dist} = d(\theta_s + \rho k_d) | (eq 3a) |
W_{dist} : “apparent” water content at saturation of distribution zone, as a depth and accounting for pesticide sorption potential to organic carbon d : thickness of distribution zone \rho : bulk density of distribution zone k_d : organic carbon distribution coefficient (product of K_{oc} and organic carbon content, see eq 18) \theta_s : volumetric saturation moisture content (=porosity) |
|
| C^w_0 = \frac{M^{app}_0}{W_{dist}} | (eq 3b) |
C^w_0 : initial dissolved pesticide concentration (at application) in distribution zone, after accounting for sorption potential by organic carbon M^{app}_0 : amount of pesticide applied in mass per unit area |
|
| C^D(t) = C^D_0 exp(-at) | (eq 4) |
C^D(t) : total pesticide concentration in water and on soil in distribution zone, at time t (as measurable in a soil sample) C^D_0 : total pesticide concentration in water and soil of distribution zone, at time 0 |
|
| a = \frac{0.69}{t_{1/2}} | (eq 5a) |
| a : First order degradation rate, as an alternate way of expressing the half life | |
| C^D_0 = \frac{M^{app}_0}{d} | (eq 5b) |
| C^w_{prfl} = C^w_0 exp(-(\frac{q}{W_{dist}} + a)t ) | (eq 6) |
| C^w_{prfl} : Dissolved concentration of pesticide in water in preferential flow path | |
| C^w_{prfl} = C^w_0 exp(-a t) | (eq 7) |
| C^w_{gr} = \beta C^w_{prfl} | (eq 8) |
C^w_{gr} : concentration of pesticide in shallow groundwater after dilution by pesticide-free recharge through soil matrix \beta : fraction of total recharge that reaches water table via preferential flow paths; this reflects matrix flow as well as preferential flow through areas where pesticide is not applied |
|
| I = \frac{C^w_{lim}}{C^w_{gr}} | (eq 9) |
I : “Groundwater pesticide index”, simply the ratio of the predicted concentration in shallow groundwater to a target maximum concentration C^w_{lim} : Reference value of pesticide concentration in water, such as an analytical detection limit |
|
| I = \frac{C^w_{lim}}{\beta C^w_0} exp(at) | (eq 10) |
| \Phi = \frac{M^{app}_0}{C^w_{lim}d} = \frac{C^w_0 (\theta_s + \rho k_d)}{C^w_{lim}} | (eq 11) |
| \Phi : fraction that the initial pesticide concentration in soil needs to be reduced to reach the limiting concentration in groundwater | |
| C^w_{lim} = \frac{(\theta_s + \rho k_d)C^w_0 }{\Phi} | (eq 12) |
| I = \frac{\rho k_d exp(at)}{\beta \Phi} | (eq 13) |
| ln(I) = \frac{0.69}{t_{1/2}}t + ln(\frac{\rho k_d}{\beta \Phi}) | (eq 14) |
| 0.69\frac{t}{t_{1/2}} + (ln(\rho k_d) - ln(\beta \Phi)) = 0 | (eq 15) |
| t_{1/2} (ln(\frac{\beta \Phi}{\rho}) - ln(k_d)) = 0.69t | (eq 16) |
| TGUS_e = t_{1/2} (ln(\frac{\beta \Phi}{\rho}) - ln(k_d)) | (eq 17a) |
| TGUS_e : Theoretical GUS on a natural log basis | |
| leachers: TGUS_e > 0.69t Note: classification as a “leacher” means that a pesticide reaches groundwater in a concentration at or above the analytical detection limit. |
(eq 17b) |
| non-leachers: TGUS_e < 0.69t |
(eq 17c) |
| k_d = f_{oc} K{oc} | (eq 18) |
k_d : sorption coefficient f_{oc} : fraction of soil mass that consists of organic carbon |
|
| TGUS_e = t_{1/2} (ln(\xi) - ln(K_{oc})) | (eq 19a) |
| \xi = \frac{\beta \Phi}{f_{oc} \rho} | (eq 19b) |
| \xi : constant combining two variable factors of a site and a pesticide. The denominator represents the strength of sorbing organic carbon in the distribution zone, which tends to hold back an applied pesticide from escape into the transmission zone. The numerator represents the ratio of initial pesticide input to detection limit. This corresponds to the 4 factor in the GUS equation 1a. | |
| TGUS = t_{1/2}(log(\xi) - log(K_{oc}) | (eq 20a) |
| TGUS : TGUS on a log 10 basis for direct comparison to GUS | |
| leachers: TGUS > 0.30t | (eq 20b) |
| non-leachers: TGUS < 0.30t | (eq 20c) |
| \frac{\beta \Phi}{f_{oc} \rho} = 10^{3.4} = 2240 | (eq 21) |
| GUS: log(t_{1/2})*(4 - log(K_{oc})) = 1.8 | (eq 22) |
| TGUS: t_{1/2}*(3.4 - log(K_{oc})) = 0.30*t | (eq 23) |
| \widetilde{log(t_{1/2})} = \frac{log(\bar{t_{1/2}})}{\bar{t_{1/2}}} t_{1/2} | (eq 24) |
| \frac{log(\bar{t_{1/2}})}{\bar{t_{1/2}}} = \frac{log(76)}{76} = \frac{1.88}{76} = 0.025 | (eq 25) |
| TGUS* = \widetilde{log(\bar{t_{1/2}})}*(3.4 - log(K_{oc})) - 0.0075 t | (eq 26a) |
| leachers: TGUS* > 0 | (eq 26b) |
| nonleachers: TGUS* < 0 | (eq 26c) |
| TGUS* = \widetilde{log(t_{1/2})}*(3.4 - log(K_{oc})) - 1.8 | (eq 27) |
| TGUS = t_{1/2}(log(\xi) - log(K_{oc}) - 0.30t | (eq 28a) |
| leachers: TGUS > 0 | (eq 28b) |
| non-leachers: TGUS < 0 | (eq 28c) |
| T_{LRP}= 3.33 t_{1/2}(log(\xi) - log(K_{oc})) for K_{oc} < \xi | (eq 29) |
| T_{LRP} = 0 for K_{oc} >= \xi | (eq 29b) |
The parameters are:
K_{oc} : organic carbon partition coefficient for pesticide
t_{1/2} : degradation half life of pesticide in soil
M^{app}_0 : pesticide application rate, mass per unit area
d : thickness of distribution zone
\beta : fraction of treated area that consists of preferential flow paths
C^w_{lim} : analytical detection limit
\rho : bulk density of distribution zone
f_{oc} : fraction organic carbon of distribution zone
References
(Add some Tammo or others’ papers about preferential solute transport in soil)
Footnotes
Tammo S. Steenhuis, Naaran Brindt, Steven Pacenka, Brian K. Richards, J-Yves Parlange, Bahareh B. Hassanpour. A theoretical underpinning of the pesticide Groundwater Ubiquity Score (GUS) J. Hydrol. Hydromech., 72, 2024. URL: https://doi.org/10.2478/johh-2024-0016↩︎
Gustafson, D.I., 1989. Groundwater ubiquity score: A simple method for assessing pesticide leachability. Environmental Toxicology and Chemistry, 8, 4, 339–357. https://doi.org/10.1002/etc.5620080411↩︎
Steenhuis et al. 2024.↩︎