Title: Applied climatology vs. applied meteorology
1Applied climatology vs. applied
meteorology
- From the AMS glossary
- applied meteorologyA field of study where
weather data, analyses, and forecasts are put to
practical use. Examples of applications include
environmental health, weather modification, air
pollution meteorology, agricultural and forest
meteorology, transportation, value-added product
development and display, and all aspects of
industrial meteorology. - applied climatologyThe scientific analysis of
climatic data in the light of a useful
application for an operational purpose.
Operational is interpreted as any specialized
endeavor within such as industrial,
manufacturing, agricultural, or technological
pursuits This is the general term for all such
work and includes agricultural climatology,
aviation climatology, bioclimatology, industrial
climatology, and others.
2Changnon (1995) diagram of applied climatology
3Climate Data and Variables
4Primary data collection
- Primary data collected via relatively cheap data
loggers or transmitted wirelessly
5Secondary data collection
6Data Sources
- WMO
- http//www.wmo.int/pages/index_en.html
- NOAA
- http//www.noaa.gov/
- National Weather Service
- http//www.weather.gov/
- National Climatic Data Center
- http//www.ncdc.noaa.gov/oa/ncdc.html
- Earth Systems Research Lab
- http//esrl.noaa.gov/psd/products/analysis/
- State Climatology offices
- http//nsstc.uah.edu/aosc/
- Regional Climate centers
- http//www.sercc.com/
7NCDC
- U.S. Stations
- http//lwf.ncdc.noaa.gov/oa/climate/stationlocator
.html - Climatological Data
- http//www7.ncdc.noaa.gov/IPS/cd/cd.html
- Monthly summary by state for all stations
- Local Climatological Data
- http//www7.ncdc.noaa.gov/IPS/lcd/lcd.html
- Monthly summary for individual stations
8NCDC Climate Divisions
Divisional means and anomalies since 1895
for Temp,Precip,PDSI
9Questions about observations and data
- Is the instrument calibrated properly? (accuracy)
- Is the instrument recording representative data?
(validity) - Spatial anomalies?
- What is the potential for bias?
- Is the instrument properly sited?
- Is the instrument recording too coarse data?
(precision) - How are observations interpolated?
- Is the data appropriate for your research
purposes?
10Ideal siting
- Open location with low vegetation
- Horizontal distance of 2 x vertical height of
nearest object - No nearby artificial heat sources
- Not in unusual microclimate
- Anemometer at 10 m elevation
- Other instruments at 1.5-2 m elevation
11Siting variability
- Orland, CA
- Marysville, CA
- (surfacestations.org)
12Issues over time
- Stations move
- Surroundings change
- Instrumentation change
- Observation changes
- Time
- Frequency
13Time of observation bias
- 24-hour observations taken at
- Midnight (all first-order stations)
- Early morning (6am-8am) especially farm
stations - Evening (6pm-10pm)
14Types of stations
- First-order station measures primary weather
variables more or less continuously, reporting
hourly (at least) - Second-order station same as first-order, though
usually less than 24 hour coverage - Cooperative station usually takes observations
one time per day
15Automated Surface Observation System (ASOS)
- Debuted in US in 1990s
- Controls all first-order stations presently
16ASOS first-order stations
- Report hourly values
- Report sub-hourly only if conditions
significantly change - Report maximum/minimum temperature every six
hours and every day - Are geared towards aviation purposes
17Things ASOS measures
- YES
- Clouds on vertical to 12,000
- Surface visibility and obstructions
- Present weather
- Temperature / dew point
- Pressure / altimeter
- Wind
- Precipitation accumulation
- Significant weather changes
- NO
- Clouds off-vertical or above 12,000
- Variable visibility
- Mixed precipitation
- Lightning
- Tornado
- Snowfall
- Snow on ground
18Coop Stations
19Climate Variables
- Temperature
- Actual vs Apparent
- Precipitation
- Measurement
- Gauge
- Radar
- Satellite
- Daily, hourly, sub-hourly
- Snow/frozen
- Dew point/humidity
- Cloud cover
- Wind direction/speed
- Pressure
- Lightning/thunderstorm days
- Sunshine/radiation
- Pan Evaporation
- Soil moisture/temperature
- Upper level sounding
- SST
20Temperature measurement
Other methods?
Stevenson screen/cotton shelter
21Precipitation measurement
Weighing gauge (NOAA)
Tipping bucket (Wikipedia)
Standard gauge (Wikipedia)
Radar (NOAA)
22Radar estimates of precipitation
- Produced in 1 hour and storm total maps
- hail and sleet may reduce accuracy
- Eastern US Radar estimates corrected by ground
observations - Western US Long-term climatological
interpolations done
23Dewpoint climatology (PRISM)
24Cloud Cover Climatology
25(No Transcript)
26January ws/wd climatology
27Thundarr Days
28Sea surface temperatures
Source JHUAPL
29Pan evaporation / lysimiter
USDA
30Upper air observations
- Radiosonde
- Developed in 1928 flourished since WW2
- Temperature, humidity, pressure
- Rawinsonde
- Similar, though provides wind speed as well
- Wind profilers
- Measure from ground
31Upper air observation locations
32Storm Data / Storm Reports
- Drought
- Dust storm
- Flood
- Fog
- Hail
- Hurricane
- Lightning
- Ocean surf
- Precipitation
- Snow / Ice
- Temperature extremes
- Tornado
- Wildfire
- Wind
33NLDN
- Detect electrical discharge through several
sensors - Triangulate location and polarity
34Derived Variables
- HDD, CDD, GDD
- Drought Indices
- http//www.drought.unl.edu/whatis/indices.htm
- SPI, PDSI, PHDI, CMI,
- Air Mass Types
- Reanalysis Data
35HDD
36Reanalysis data
- Combination of weather forecast model
initialization and analysis, and short-term
forecast - Project started in 1990s to reproduce synoptic
maps back to 1948 extrapolation to 1908 coming
soon - Two significant programs
- NCEP / NCAR NNR (USA)
- ECMWF ERA (European Union)
37Reanalysis fields produced
- Class A the most reliable class of variables
"analysis variable is strongly influenced by
observed data" - Class B the next most reliable class of
variables "although some observational data
directly affect the value of the variable, the
model also has a very strong influence on the
output values." - Class C the least reliable class of variables
NO observations directly affect the variable and
it is derived solely from the model computations
forced by the model's data assimilation process,
not by any real data. - Class D a mean field that is obtained from
climatological values and does not depend on the
model
38Reanalysis examples
39US Climate Reference Network
- Set up since 2000 to serve as reference point for
long-term climate records
40US Historical Climate Network
- Derived from previously observed data
- Many statistical routines run to attempt to
homogenize datasets
41Meteorological Assimilation Data Ingest System
(MADIS)
42Levels of aggregation
43(No Transcript)
44Levels of aggregation
45Levels of aggregation
46(No Transcript)
47Levels of aggregation
48(No Transcript)