Name
Input value
Explanation
Model
Identifies the name to be assigned to the model being verified. This text string is written out in the first column of the ASCII output so that verification statistics from multiple models may be differentiated.
Grid Resolution
The nominal spacing for each grid square in kilometers. The variable is not used directly in the code, but subsequent variables in the configuration file are defined in terms of it. Therefore, setting this appropriately will help ensure that appropriate default values are used for these variables.
Forecast Field
Specifies the model variable and the corresponding vertical level to be verified. The GRIB code itself or the corresponding abbreviation may be used to specify which model field is to be verified. A slash and a level indicator in the form ANNN, ZNNN, or PNNN must follow each GRIB code. The level indicators correspond to an accumulation interval (A), a single vertical level (Z), or a single pressure level (P). NNN indicates the accumulation or level value.
Obs Field
Specifies the Obervation variable. See above.
Forecast Convolution Radius
Defines the radius of the circular convolution applied to smooth the raw fields. The radius is specified in terms of grid units.
Obs Convolution Radius
See above.
Forecast Convolution Threshold
Specifies the threshold values to be applied to the convolved field to define objects. The thresholds are specified using the Fortran conventions of gt, ge, eq, ne, lt, le.
Obs Convolution Threshold
See above.
Forecast Merge Flag
No Fcst Merging
Double Threshold Only
Fuzzy Engine Only
Both Methods
The fcst_merge_flag and obs_merge_flag variable control what type of merging techniques will be applied to the objects defined in each field.
Obs Merge Flag
No Obs Merging
Double Threshold Only
Fuzzy Engine Only
Both Methods
Match Flag
No Matching
Additional Merging in Both Fields
Additional Merging in Fcst Field
Matching With No Additional Merging
The match_flag variable controls how matching will be performed when comparing objects from the forecast field to objects from the observation field. An interest value is computed for each possible pair of forecast/observation objects. The interest values are then thresholded to define which objects match. If two objects in one field happen to match the same object in the other field, then those two objects could be merged. The match_flag controls what type of merging is allowed in this context.
Total Interest Threshold
Should be set between 0 and 1. This threshold is applied to the total interest values computed for each pair of objects. Object pairs that have an interest value that is above this threshold will be matched, while those with an interest value that is below this threshold will remain unmatched. Increasing the threshold will decrease the number of matches while decreasing the threshold will increase the number of matches.
Name
Input value
Explanation
Mask Missing Flag
No Masking
Mask Fcst with Obs
Mask Obs with Fcst
Mask Both
Specifies how missing data in the raw model and observation fields will be treated.
Mask Grid
Specifies a pre-defined NCEP grid with which to mask the raw forecast and observation fields. The predefined grids are specified as GNNN where NNN is the three digit designation for the grid. By default, no masking grid is applied.
Mask Grid Flag
Do Not Apply
Fcst Only
Obs Only
Both Fcst and Obs
Specifies how the mask_grid should be applied. By default, the masking grid is not applied.
Mask Polyline
Specifies the name of a file that defines a lat/lon polygon to be used in masking the raw forecast and observation fields. Several masking polygons used by NCEP are predefined in the data/poly subdirectory of the MET distribution. Users can easily define and apply new masking polygons similar to the predefined ones. By default, no masking polygons are used. Similar to the mask_grid variable above.
Mask Polyline Flag
Do Not Apply
Fcst Only
Obs Only
Both Fcst and Obs
Specifies how the masking polygon should be applied. Similar to the mask_grid_flag variable above.
Forecast Raw Threshold
Used to threshold the raw fcst field. Thresholds are specified using the Fortran conventions of gt, ge, eq, ne, lt, le to indicate greater than, greater than or equal to, equal to, not equal to, less than, and less than or equal to, respectively. By default, the raw fields are thresholded greater than or equal to zero.
Obs Raw Threshold
See above.
Bad Data Threshold
Must be between 0 and 1. When performing the circular convolution step if the proportion of bad data values in the convolution area is greater than or equal to this threshold, the resulting convolved value will be set to "bad data". If the proportion is less than this threshold, the convolution will be performed on only the valid data.
Forecast Area Threshold
Specifies the area threshold values to be applied to the defined objects.
Obs Area Threshold
See above.
Forecast Intensity Percent
The fcst_inten_perc, fcst_inten_perc_threshold, obs_inten_perc, and obs_inten_perc_threshold variables specify the intensity threshold values to be applied to the defined objects. For each object defined, the intensity values within the object are sorted, and the requested intensity percentile value is computed. By default, the maximum value is computed since the intensity percentiles are set to 100. Any objects with an intensity percentile that does not meet the corresponding intensity percentile threshold specified will be discarded.
Forecast Intensity Percent Threshold
Obs Intensity Percent
Obs Intensity Percent Threshold
Forecast Merge Threshold
The fcst_merge_threshold and obs_merge_threshold variables are used to define larger objects for use in merging the original objects. These variables define the threshold value used in the double thresholding merging technique. Note that in order to use this merging technique, it must be requested using the fcst_merge_flag and obs_merge_flag. These thresholds should be chosen to define larger objects which fully contain the originally defined objects.
Obs Merge Threshold