Climate Data Analysis

Climate Data Analysis

John GrossNPS I&M Program

GIS / Data Management Conference 3 April 2008

National Park Service

Parks and climate change

National Park Service

• Relatively undisturbed• Protected for the future• Broad geographic distribution• Large environmental gradients• Many observations

People really care about Parks

IPCC 2007 Working Group 4 FAQ 3.1.

Projected temperature change

"For us the discussion around climate change is not just a theory; it is a very stark and harsh reality.”

Patricia Cochran – Inuit Circumpolar Conference

Observation (T)

Station summary of day (SOD)

Station month summary

Atmospheric Index – SOI, PDO

Climate indices (PDSI)

Climate Divisions

Spatial Scale

TimeInvestigator-defined area of interest

minutes

Months to years

points regions

“Climate tells you what clothes to buy …

weather tells you what to wear today!”

• Climate is determined by the properties of the Earth system that define the normal range of variation in observation.

• Weather responds sensitively to local conditions. Weather forecasts are only useful for short periods into the future.

Observation (T)

Daily station summary (SOD)

Monthly station summary

Atmospheric Index – SOI, PDO

Many indices (PDSI)

Climate Divisions

Spatial Scale

Time(roll-your-own area of interest)

minutes

Months to years

points regions

Individual station data is essential

This is the basis of our understanding and extrapolation

For long-term value, QA/QC and metadata are CRITICAL

Communicating climate

• Consider the reference period– 30 year ‘normals’ (often 1971-2000)– Life of station– Since 1895

http://www.wcc.nrcs.usda.gov/snow/

Climates always change – use a thoughtful reference period

• Highly subject to:– Micrometeorological phenomena– Instrument issues – calibration, changes, etc.– Station relocation and site-specific effects

• Rigorous QA/QC difficult and time-consuming – Missing values difficult to handle and accomodate– McEachern CHIS report (2007) good example

• Metadata is essential– Almost surely need data from an established network

Individual Station Data and Climate Evaluation

Observation (T)

Daily station summary (SOD)

Monthly station summary

Atmospheric Index – SOI, PDO

Many indices (PDSI)

Climate Divisions

Spatial Scale

Time(roll-your-own area of interest)

minutes

Months to years

points regions

Describing climate of an area

• Exemplar or ‘indicator’ station(s)

• Relatively easy, once stations are selected

• Station data – local effects, reliant on single sensor, single-station bias

Multiple-station index or aggregation• Broader inference

• Strength in numbers• More complex analysis problem

California Climate Tracker - WRCC

Describing climate of an area

NCDC Climate Divisions – To the rescue?

http://www7.ncdc.noaa.gov/CDO/CDODivisionalSelect.jsp

• Monthly data from 1895 - present

• Temperature, precipitation, heating/cooling degree days

• Palmer drought index and related

• Standardized precipitation indices

• Simple data output to text file or (limited) plots

Mapped by climate division

With apologies to Hawaii and Alaska …

Temperatures in Northwest Wisconsin

Apostle Islands National Lakeshore Established

From http://www.wrcc.dri.edu/spi/divplot1map.htmlRed line – 12 month averageBlue line – 10 year running mean

1970 19901950 201019301910

Correlation in precipitation between station and climate division Jan-Mar.

(K. Wolter and D. Allured 2007)

Red = GoodGreen = badBlue = Really bad!

See: www.cdc.noaa.gov/people/klaus.wolter/ClimateDivisions/

Large-scale climate drivers are important

• Climate station inventory reports• Regional Climate Centers papers (WRCC for western states)

NEON process: http://web.utk.edu/~jweltzin/SAPOZEO/Hayden090805.htm

Consider regional indices – these may better predict ecological characteristics than local

data!

• Data and analyses– NPClime and climate inventory reports– Climate learning modules – www.meted.ucar.edu– ROMN protocol (in draft) for miscellaneous data handling – National Climatic Data Center (NCDC) – many products– Gridded data – PRISM– NRCS SnoTel, National Drought Mitigation Center, USGS

stream flow

Selected Resources and Activities