Station Quality Control Status
A key component of the CDMP 19th Century Forts Project was the quality control of digitized daily weather observations from the U.S. A series of computerized tests were conducted to enhance and ensure the accuracy of the CDMP Forts data (see the Quality Control Methodology below). If any outliers were detected through these tests, the data were manually reviewed and corrected.
Only those weather stations that have undergone all quality control tests are available. To view a list of keyed stations sorted by state, availability, or the completeness of quality control tests, select one of the "Go" buttons below. You can also access static maps for stations indicating the location of quality-controlled vs. non-quality-controlled stations.
Sort by state:
Go
Sort by Availability:
Go
Sort by First 3 Tests Completed:
Go
Quality Control Methodology
The first group of tests focused on identifying significant errors in the entire set of keyed data and inconsistencies between it and its corresponding metadata. The primary functions of these three tests were as follows:
- Reviewed each data type and identified logically or physically impossible values. For example, a non-existent cloud-type abbreviation would be a logical impossibility, while a relative humidity reading of 0% would be a physical impossibility.
- Verified the consistency between the date of the keyed data and the date specified in the metadata.
- Conducted a detailed comparison of the keyed data against the metadata element-by-element basis. Any elements that were either requested in the metadata but not keyed or keyed but not requested will be flagged.
The second set of tests examined the accuracy of temperature, precipitation, and snowfall measurements. For each variable, we calculated the monthly sums and means, which were then compared to the values recorded by the observer. Significant discrepancies were flagged for further review. Additionally, we assessed subjectively determined threshold cut-offs for extreme values.
- Monthly temperature tests included comparisons of various elements such as mean, maximum, minimum, and hourly temperatures and assessments of the temperature range and both dry bulb and wet bulb temperatures. Daily tests examined climatologically extreme values, daily spikes, and unusually large or small diurnal temperature ranges. Additionally, maximum, minimum, and hourly temperature comparisons were conducted to identify inconsistencies.
- Monthly precipitation assessments involved manually reviewing the five highest monthly precipitation totals. Daily assessments included checking climatological extreme values and verifying that the sum of hourly and daily measurements matches the reported daily and monthly totals. If the sum of the daily values equaled the monthly total reported by the observer, the remaining days were filled with zeros to finalize the precipitation data.
- The five largest monthly snowfall amounts and snow depths were manually evaluated. Comparisons were made between the total precipitation and the accumulated snow, as well as between the monthly totals of snowfall and snow depth. Additionally, snowfall was compared against the average minimum monthly temperatures.
For further details on quality control of the 19th century weather data, see this report (pdf).