Concentration models tiers¶

In addition to the possibility for users to work with their own choices for all settings, MCRA implements four tiers from two documents:

The optimistic and pessimistic basic assessments from the EFSA 2012 Guidance on the Use of Probabilistic Methodology for Modelling Dietary Exposure to Pesticide Residues [3].
Tier 1 and 2 from the European Commission working document SANTE-2015-10216 rev. 7 (2018) on risk management aspects related to the assessment of cumulative exposure [37].

Overview¶

Table 34 Tier overview for module Concentration models.¶
Name	EFSA 2012 Optimistic	EFSA 2012 Pessimistic	EC 2018 Tier 1	EC 2018 Tier 2
Default concentration model	Empirical	NonDetectSpikeLogNormal	Empirical	Empirical
Include MRL fallback model	false	true	false	false
Restrict LOR imputation to authorised uses		false	false	false
Non-detects replacement	ReplaceByZero	ReplaceByLOR	ReplaceByLOR	ReplaceByLOR
Factor f (f x LOR)		1	0.5	0.5
MRL Factor (f x MRL)		1
Sample based	true	true	true	true
Imputation of missing values	false	true	true	true
Correlate imputed values with sample potency	false	true	true	false
Use occurrence patterns for imputation			true	true
Parametric uncertainty	false	true	false	false

The sections below describe the settings specified by each tier in detail.

EFSA 2012 Optimistic¶

Use the optimistic model settings according to the EFSA Guidance 2012. Non-detects and missing values are replaced by zero.

Table 35 Tier definition for EFSA 2012 Optimistic.¶
Name	Setting
Default concentration model	Empirical
Include MRL fallback model	false
Non-detects replacement	ReplaceByZero
Sample based	true
Imputation of missing values	false
Correlate imputed values with sample potency	false
Parametric uncertainty	false

EFSA 2012 Pessimistic¶

Use the pessimistic model settings according to the EFSA Guidance 2012. A nondetect spike lognormal model is fitted to the positive residue values. Non-detects are replaced by the LOR. When the number of positives is smaller than 2, the maximum residue limit (if available) is used instead. Missing values are imputed.

Table 36 Tier definition for EFSA 2012 Pessimistic.¶
Name	Setting
Default concentration model	NonDetectSpikeLogNormal
Include MRL fallback model	true
Restrict LOR imputation to authorised uses	false
Non-detects replacement	ReplaceByLOR
Factor f (f x LOR)	1
MRL Factor (f x MRL)	1
Sample based	true
Imputation of missing values	true
Correlate imputed values with sample potency	true
Parametric uncertainty	true

EC 2018 Tier 1¶

Table 37 Tier definition for EC 2018 Tier 1.¶
Name	Setting
Default concentration model	Empirical
Include MRL fallback model	false
Restrict LOR imputation to authorised uses	false
Non-detects replacement	ReplaceByLOR
Factor f (f x LOR)	0.5
Sample based	true
Imputation of missing values	true
Correlate imputed values with sample potency	true
Use occurrence patterns for imputation	true
Parametric uncertainty	false

Input tiers¶

Table 38 Input tiers for EC 2018 Tier 1.¶
Module	Input tier
Occurrence patterns	EC 2018 Tier 1
Concentrations	EC 2018 Tier 1

EC 2018 Tier 2¶

Table 39 Tier definition for EC 2018 Tier 2.¶
Name	Setting
Default concentration model	Empirical
Include MRL fallback model	false
Restrict LOR imputation to authorised uses	false
Non-detects replacement	ReplaceByLOR
Factor f (f x LOR)	0.5
Sample based	true
Imputation of missing values	true
Correlate imputed values with sample potency	false
Use occurrence patterns for imputation	true
Parametric uncertainty	false

Input tiers¶

Table 40 Input tiers for EC 2018 Tier 2.¶
Module	Input tier
Occurrence patterns	EC 2018 Tier 2
Concentrations	EC 2018 Tier 2