Forecaster Performance Criterion 4.2

Impact of unrepresentative or erroneous observations

Performance Criterion Description:

The forecaster needs to be able to discern and discount or consider appropriately unrepresentative or erroneous observations discovered during the analysis/diagnosis/forecast process.    

Scenario:       

During an assessment period the forecaster is given a flawed set of data for the forecast region. The forecaster needs to recognize and reject the erroneous data when producing the forecast.   

Evidence:

During the analysis/diagnosis/forecast process the forecaster

• notes suspect observations that  are unrepresentative of surroundings
• discounts erroneous data
• knows the most frequent errors that might occur.
• applies the proper verification methods for the available data
• evaluates the impact of the erroneous observational data on the operational work
• errors recognized and discounted
• suspect observations compared to surroundings
• compliance/noncompliance with conceptual models considered
• corrected information obtained
• communicated findings