flag and remove 0's from most weighted average calculations

[dylanbeaudette]

Apart from rock fragment volume, there aren't many cases were a 0 in the horizon or component data makes sense. These values should be:

• flagged for follow-up (future) MLRA projects
• filtered from weighted mean calculations because of the skewing effect on the final values

Filtering will be required at the horizon and component levels.

Horizon properties:

Component properties:

[dylanbeaudette]

There are some 0's that must be retained: CaCO3 and EC, and SAR.

[dylanbeaudette]

Investigating om_r values equal to 0. There are very few cases where 0% organic matter is possible.

Fortunately, there are only 63 records with A horizons with 0% organic matter.

select component.areasymbol, compkind, compname, hzname, om_r from ssurgo.component JOIN ssurgo.chorizon USING (cokey) where om_r = 0 AND hzname ~~ 'A'

A couple of examples:

 areasymbol |      compkind      |     compname      | hzname | om_r
------------+--------------------+-------------------+--------+------
 mt033      | Series             | Yamacall          | A      |    0
 mt649      | Series             | Yamacall          | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn625      | Series             | Ratroot           | A      |    0
 mn099      | Series             | Kalmarville       | A      |    0
 mn099      | Series             | Kalmarville       | A      |    0
 mn099      | Series             | Kalmarville       | A      |    0
 mn099      | Series             | Kalmarville       | A      |    0
 mt055      | Series             | Yamacall          | A      |    0
 mt065      | Series             | Yamacall          | A      |    0
 nm692      | Series             | Atchee            | A      |    0
 mt025      | Series             | Yamacall          | A      |    0
 ut629      | Series             | Reef              | A      |    0
 mn613      | Series             | Ratroot           | A      |    0
 mn613      | Series             | Ratroot           | A      |    0
 sc051      | Series             | Newhan            | A      |    0
 ut642      | Series             | Radnik            | A      |    0
 ut643      | Family             | Monue family      | A      |    0
 ut643      | Series             | Needle            | A      |    0
 ca668      | Series             | Whitewolf         | A      |    0
 ut685      | Series             | Reef              | A      |    0
 ut685      | Taxadjunct         | Retsabal          | A      |    0
 ca805      | Taxon above family | Typic Haplosalids | A      |    0
 ca666      | Series             | Lokern            | A      |    0

[dylanbeaudette]

14 addressed the more important issue. 0% organic matter is a data quality issue that cannot be resolved here.