Weather 2000 provides a wide spectrum of historical weather data
resources to our clients. Our experienced climatologists and
meteorologists have worked for the National Weather Service (NWS) and
the National Climatic Data Center (NCDC), possessing a comprehensive
knowledge of government weather records and observational procedure.
By understanding what to acquire we can help you with the following issues
- Weather station locations
- Missing, Estimated and Interpolated data
- Data resolution - daily, monthly, seasonal and annual data
- Weather variables - degree days, temperature, rainfall,snowfall, etc.
- Familiarization with weather station and instrumentation idiosyncrasies:
- Station type (1st order and 3rd order [coop])
- observational period of record
- observational status (time and frequency of observations)
- variable availability
- micro-climate trends
- instrumentation biases
- station location changes
- recording errors
- more info: Weather Data FAQ
|
Government quality control:
Weather observations in the U.S are taken at least daily, but
occasionally these data may be missing or contain human, instrumentation
or computer errors. NCDC compiles, filters and quality-controls NWS observations, making the data "Official". (see Official data below)
Additional data "cleaning" and adjusting:
Recently there has been a lot of attention focused on the idea of
"cleaning" weather data-sets. Many weather stations have experienced
temperature trends and alterations in their surroundings and
micro-environments throughout their history. However, this should all
be considered within the realm of climate variability. In most
instances, data outliers and extreme outliers should not be adjusted
or excluded when conducting statistical analyses on historical weather
data-sets. All data points play an essential role when determining
potential climate outcomes and when calculating standard deviations.
|
|
 |
Understanding that the margin of confidence in a seasonal forecast can be 100's of degree days, is it
really essential to clean/normalize the historical data, or is it better to primarily make subjective adjustments to the forecast? |
 |
Taken in this context, a fraction of a degree [as caused by daily data cleaning and adjustments] will not make
a significant difference to a monthly temperature average or to seasonal degree day calculations. Rather, a subjective adjustment to the
forecast should probably be made. Subjectivity has an important place in long-range seasonal forecasting.
- Accredited to Dr. Daniel S. Wilks, Professor of Statistical Meteorology, Cornell University.
|
|
Note:
For much more technical scientific analysis (such as determining
regional warming of urban versus rural areas), adjusting data-sets with
academically proven techniques can be useful. For such matters we
highly recommend referring to ongoing research conducted by NCDC:
|
|
|
Official data:
Since there is no such thing as "real-time official" data, for recent
information you will have to rely on "preliminary" data. For Official
data, NCDC creates Edited LCD publications
which are now available via the Internet as quickly as 3 weeks after the
observation is made.
It should be noted that although other services may provide their own
quality-controlled data, only NCDC has the legal authority to produce
and archive Official U.S. weather records. Thus, for contract
settlements and legal disputes, only Official data from NCDC are valid
and should be utilized.
|
|
|
Data costs:
Fortunately in the U.S., weather data is widely available and extremely
inexpensive compared to other countries. Weather 2000 will often provide
historical data as part of their forecasting services.
Even NCDC produces a wide variety of printed and Internet-accessible historical data products
for free or at minimal cost. Either way, the bottom line is you should
never pay large fees for U.S. weather data, whether it's "adjusted" or not.
Understand that high-priced retailed data-sets by private firms are
unfair manipulations of businesses and consumers.
|
