The Multiple Uses of a Comprehensive Climatology Database

1
The Multiple Uses of a Comprehensive Climatology
Database
John P. Gagan, Steve A. Listemaa and Eric E.
Carpenter NWS WFO Jackson, MS
And
Jennifer L. Colson
NWS WFO Tampa Bay, FL
2
The Purpose of this Presentation

• List useful climate information readily available
  to all National Weather Service Forecast Offices
  (NWS WFOs)
• Identify methods of data collection
• Compare methods of storing and accessing the data
• Examine examples of how these databases can
  improve forecasts and services

3
Why take the time and effort to create these
databases?

• The NWS is mandated to protect life and property
• In order to protect life and property, the NWS
  forecaster is responsible for forecasting extreme
  (anomalous) events as far in advance as possible
• These databases provide a comprehensive knowledge
  of a CWAs weather history and offer an explicit
  investigation into what can occur

4
Climatological Data

• A portion of the vast NCDC archive includes
  hourly and daily surface and upper air data for
  thousands of stations across the United States
  for a variety of time spans
• This large data set serves as a footprint for
  each NWS WFO county warning area (CWA) and is an
  essential resource for all forecasters

5
Climatological Data (cont)

• Available Surface Data
• Hourly -- From station inception through the
  present
• Wind Speed, Direction and Gusts
• Visibility
• Temperature
• Dewpoint
• Ceiling Height and Coverage
• Pressure
• Rainfall

6
Climatological Data (cont)

• Available Surface Data (cont)
• Daily (Varying Time Spans)
• High/Low Temperature
• 24-hour Rainfall
• 24-hour Snowfall
• Snow Depth
• Upper Air
• Twice Daily, with intermediates

7
Data Acquisition

• Climatological data sets available through the
  National Climatic Data Center
• www.ncdc.noaa.gov
• Most data available on CDROMs or web-based
  queries
• Many of these data sets available in digital
  ASCII format
• Lists of station histories provided via ftp (also
  in Adobe Acrobat, Word or Text files)

8
Data Acquisition (cont)

• Available data for the Jackson, MS CWA
• Hundreds of daily coop reports, with histories
  dating back as far as 1890s
• Eleven hourly surface observation sites with
  histories as far back as 1901 (most back to the
  1970s)
• All available North American RAOB data from 1946
  onward (best coverage from 1957 through the
  present)

9
Data Manipulation

• Files acquired through CDROM and download were
  formatted to comma delimited (if not already in
  this format)
• Upper air mandatory-level data from the KJAN site
  extracted using GEMPAK
• Comma delimited data saved as text files

10
Data Manipulation (cont)

• Data delimited by Year, Month, Day and Hour (when
  applicable)
• Once all the acquired files have been saved as
  text files they are imported to a MySQL database
• Databases are created for each data type

11
KJAN Daily Database Query The Old Database

• Initially, Microsoft Access was used as it
  allowed for easy access to all data
• Queries were created to search for any year,
  month, date, parameter, or combination therein
• After testing the database, we found that the
  Access GUI was user-unfriendly
• A simple, web-based version received a much
  better response from users

12
KJAN Daily Database Query The Old Database
Example
How many times has KJAN observed low temperatures
at or below 32 F in October?
13
KJAN Daily Database Query Example

• Between 1930 and 2002 (2232 observation records)
  a temperature of 32 F or lower was observed only
  30 times in October
• Obviously a 98th percentile event!

14
The New KJAN Daily Database

• COOP Search
• Search for Tmax, Tmin, and Precip for a station

15
The New KJAN Daily Database

• Web Table output
• Simple
• Basic Info Fast

16
The New KJAN Daily Database

• Text output
• Save as a .txt
• Easily imported into a spreadsheet
• Quick analysis

17
The New KJAN Daily Database

• METAR Search
• Example just temperature

18
The New KJAN Daily Database
19
The New KJAN Daily Database
Graphing Feature
20
The New KJAN Daily Database
21
Special Queries

• Show all observations with Temps lt 15F in the
  year 1982
• This is just a partial listing

22
KJAN Daily Database Query Example
How many times has KJAN observed low temperatures
at or below 32 F in October?
23
KJAN Daily Database Query Example

• Between 1930 and 2002 (2232 observation records)
  a temperature of 32 F or lower was observed only
  30 times in October
• Obviously a 98th percentile event!

24
KJAN Upper Air Database

• All available North American raob data (TTAA/TTBB
  format) available from NCDC converted to GEMPAK
• Each sounding viewable through NSHARP
• Mandatory level data extracted and put into text
  files
• Text data then imported to Microsoft Access

25
KJAN Upper Air Database

• Sounding coverage available in the database

26
KJAN Upper Air Database
0000 UTC 22 February 1971 KJAN Sounding Delta
Outbreak
27
KJAN Upper Air Database Text Data

• Mandatory Text data includes
• Temperature
• Dewpoint
• Heights
• Wind Direction/Speed

28
KJAN Upper Air Database Text Data (cont)
29
KJAN Upper Air Database Text Data
example

• How many times has the 850hpa temperature been
  observed at -10C or lower in January?

16 times out of 1300 observations
30
Uses of the Databases

• All databases can be linked to one another
  (Hourly to Daily to Upper Air)
• Break down any data set for any period of time,
  for any parameter
• Statistical analysis of parameters pertinent to
  general forecasting, aviation, fire weather, etc.
• Verify historical occurrences of special weather
  criteria such as wind advisories, wind chill,
  heat indices, Red Flag, etc.

31
Uses of the Databases (cont)

• Sounding data can be used to create composites of
  conditions that have led to heavy rainfall,
  severe weather and winter weather in your
  particular area
• Inclusion of teleconnection data from the CPC
  allows a user to define average temps during each
  ENSO phase, PDO phase, etc.
• The user is bound only by their imagination
• A few examples

32
Uses of the Databases (cont)

• Break down climatological hourly temperature and
  dewpoint curves to be used in the WFOs Graphical
  Forecast Editor (GFE)
• Further analysis can break down effects of cloud
  cover and wind direction on the hourly
  temperature and dewpoint curves
• These curves can significantly increase accuracy
  of point forecast matrices (PFMs)

33
Hourly Temperature Curves
34
Uses of the Databases (cont)

• Hourly Data
• Climo temperature/dewpoint curves
• Historical analysis of wind speed and direction
  (Homeland Security issues)
• Typical Ceilings and Visibilities associated with
  particular weather phenomena
• Analysis of wind chill/heat index values

35
Uses of the Databases (cont)

• Daily data
• Define average temperature and normal
  temperatures for a given time period
• Create box and whiskers diagrams to visually
  separate the 50th percentile from truly anomalous
  temperatures

36
Uses of the Databases (cont)
Craven 2003
37
Uses of the Databases (cont)
Late October KJAN stats
38
Uses of the Databases (cont)

• Excellent source for investigating significant
  weather events
• Linking Daily data with text Upper Air data gives
  a 3-D view of anomalously hot/cold air masses
• Can take model and ensemble consensus in the
  extended forecast period to explicitly quantify
  what an air mass is capable of
• Quite powerful when used in tandem with NCEP/NCAR
  reanalysis data

39
Uses of the Databases (cont)

• NCEP/NCAR Reanalysis data available at the
  Climate Diagnostics Center
• http//www.cdc.noaa.gov/cdc/reanalysis/
• Visualization of surface and upper air parameters
  for any date back to 1948
• GRADS image creation allows user to define
  viewing area
• Highly recommend this site

40
Uses of the Databases (cont)

• Lets briefly revisit our upper air database
  example

• How many times has the 850hpa temperature been
  observed at -10C or lower in January?

16 times out of 1300 observations
41
Uses of the Databases (cont)

• Where do the associated daily high and low
  temperatures fit in with respect to normal?

42
Uses of the Databases (cont)

• Using the Daily database, a distribution of all
  high and low temperature recorded in January can
  be exported to Microsoft Excel
• Finding the max, min, median, 25th percentile and
  75th percentile values for both the high and low
  temperature in January, a box and whiskers
  diagram can be created

43
Uses of the Databases (cont)
January
40
33.3
27
24
16
2
HIGH
LOW
44
Uses of the Databases (cont)

• Although not surprising by any means, a
  forecaster can quantify that 850hpa temperatures
  of ? -10 C lead to surface high and low
  temperatures in the lowest quartile of the
  temperature distribution for January
• Statistics just like these are currently being
  developed at KJAN and are inserted into a binder
  next to the synoptic forecasters workstation

45
Benefits of the Database

• Increased knowledge of your forecast area of
  concern
• Statistics can easily be generated for any
  station, parameter, and time period
• Generate smart tools for use in IFPS (GFE)
• Increased situational awareness of anomalous air
  masses/events in the extended forecast period
• Enhances in-depth studies of any weather
  phenomena that affects your region

46
Future Endeavors

• Complete the Climate Database Retrieval System
  for each WFO within Southern Region
• Each office will have quick access to
• Hourly observations and COOP observations for
  your area
• ALL raobs from North America
• This data will be queried through a simple
  web-based platform

47
Future Endeavors

• Acquisition of hourly coop rainfall data
• In depth studies of heavy rainfall events by
  season
• Define what heavy rainfall means for an area
• Acquisition of snow cover data
• Contour all upper level height data using GEMPAK
  for quick reference

48
MANY THANKS!

• Eric, Jennifer, Steve and I would like to thank
  Alan Gerard (MIC) and Jeff Craven (SOO) for
  allowing us to tackle this large project
• My wife, Jennifer, for showing me the finer
  points of Microsoft Access
• Also, the CDC, CPC and NCDC for their monumental
  efforts in collecting and creating quality data
  sets and web generated products

49
Questions, comments?
Contact myself or Eric at
John.Gagan_at_noaa.gov Eric.Carpenter_at_noaa.gov