Several validations are done to help users create and manage viable multivariate analyses.  There are a total of nine.  All are applied at run time.  The first seven are also applied as multivariate analyses are created.  When an invalid combination is detected, HEC-EFM identifies the failed validation parenthetically at the end of the corresponding message.

  1.  Current relationship has relationship-defined water year selected (all relationships invalid)
  2.  Pairing relationship has relationship-defined water year selected (pairing relationship invalid)
  3.  Flow regime inactive
  4.  Relationship inactive
  5.  Not an ecovalue summation relationship
  6.  Mismatch in number of parts
  7.  Season of pairing relationship not wholly nested in current relationship
  8.  Temporal mismatch
  9.  One or more ecovalue time series for selected pairings has values less than 0 or greater than 1

The first and second validations flag use of relationships that are incompatible with multivariate analyses.  The third and fourth check whether requisite inputs will be available for multivariate analyses.  The fifth checks whether relationships in selected pairings are ecovalue summation relationships.  The sixth checks whether the flow regimes in selected pairings have the same number of parts.  The seventh checks whether relationships in selected pairings have compatible seasons.  The eighth checks whether the periods of record for selected pairings allow for computation of multivariate analyses.  The ninth checks whether ecovalue time series for selected pairings have values less than 0 or greater than 1.

Failed validations for the first seven checks are reported via the Multivariate Analyses - Select Pairings interface and at run time.  For example, Figure 72 shows a selected pairing that failed validation 6 - Mismatch in number of parts.  The number of parts for the Aquatic Structures flow regime (20,380) is different than the number of parts for the Wetland Restoration Zone flow regime (2,940).  Validation 6 checks that the number of parts for selected pairings are equal, which is a simple check intended to flag spatial discrepancies.  Invalid selections can be saved, allowing users to continue working with their EFM projects, but will fail at run time.

Figure 72.  Failed validations are reported in the Select Pairings interface.  Red text is used to highlight problems and includes the failed validation identifier (in this image, “6”).


The eighth and ninth validations are only done at run time because the periods of record and ecovalues needed for validation are generated during the compute process.  As multivariate analyses are done last, this approach does leave the possibility that a run will fail unexpectedly late in its compute cycle.  Messages about run time validation fails (for all validations) are provided in EFM output tables and via pop-up windows that can be left open as a reference for users troubleshooting multivariate analyses (Figure 73).

Figure 73.  Validations failed at run time are reported in EFM output tables (left) and warning message windows (right).