Briefing on Pairwise Approach to Climate Data 1 NOAA’s National Climatic Data Center Automated Homogenization of Monthly Temperature Series via Pairwise Comparisons Matthew Menne and Claude Williams NOAA/National Climatic Data Center Asheville, North Carolina USA
60
Embed
Automated Homogenization of Monthly Temperature Series via Pairwise Comparisons
Automated Homogenization of Monthly Temperature Series via Pairwise Comparisons. Matthew Menne and Claude Williams NOAA/National Climatic Data Center Asheville, North Carolina USA. Outline. Motivation: The United States Historical Climatology Network (U.S. HCN) - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Briefing on Pairwise Approach toClimate Data Homogenization
1 NOAA’s National Climatic Data Center
Automated Homogenization of Monthly Temperature Series via Pairwise Comparisons
Matthew Menne and
Claude WilliamsNOAA/National Climatic Data Center
Asheville, North Carolina USA
Briefing on Pairwise Approach toClimate Data Homogenization
2 NOAA’s National Climatic Data Center
Outline
1. Motivation: The United States Historical Climatology Network (U.S. HCN)
2. Overview of the “pairwise” homogenization algorithm
3. Some examples4. Impact of inhomogeneities on U.S.
temperature trends5. A word about GHCN-Daily
Briefing on Pairwise Approach toClimate Data Homogenization
3 NOAA’s National Climatic Data Center
U.S. Climate Network
Historical Climatology Network
Cooperative Observer
Briefing on Pairwise Approach toClimate Data Homogenization
4 NOAA’s National Climatic Data Center
U.S. HCN -- Version 1Monthly Data
• 1221 stations selected to comprise the HCN in mid-1980s
• Monthly dataset originally released in 1987 • Addressed the following:
– Time of observation bias (Karl et al. 1986; Vose et al. 2003)– Station History Changes (Karl and Williams 1987)
• Optimized reference series based on station history archives
– Urbanization (Karl et al. 1988) – LiG to MMTS instrument change (Quayle et al. 1991)
Briefing on Pairwise Approach toClimate Data Homogenization
5 NOAA’s National Climatic Data Center
U.S. HCN -- Version 2Monthly Data
• 1218 stations in a re-defined network• Addresses
– Time of observation bias (Karl et al. 1986; Vose et al. 2003)
– Station history (documented) and undocumented changes (Menne and Williams, Journal of Climate, in review)
• Automated pairwise comparison of series
Briefing on Pairwise Approach toClimate Data Homogenization
6 NOAA’s National Climatic Data Center
Station Siting Example of ratings assigned by Watts based upon NOAA/NCDC criteria
Class 1 - Flat & horizontal ground. Sensors located at least 100 meters from artificial heating
Class 2 - Same as Class 1, except no artificial heating sources within 30 meters.
Class 3 - Same as Class 2, except no artificial heating sources within 10 meters.
Class 4 - Artificial heating sources <10 meters.
Class 5 - Temperature sensor located next to/above an artificial heating source
Siting Classification based upon standards
for NOAA’s U.S. Climate Reference
Networkftp://ftp.ncdc.noaa.gov/pub/data/uscrn/
documentation/program/X030FullDocumentD0.pdf
www.surfacestations.org
Briefing on Pairwise Approach toClimate Data Homogenization
7 NOAA’s National Climatic Data Center
Station Siting Example of ratings assigned by Watts based upon NOAA/NCDC criteria
Class 4 - Artificial heating sources <10 meters.
Class 5 - Temperature sensor located next to/above an artificial heating source
Siting Classification based upon standards for NOAA’s U.S. Climate Reference Network
Briefing on Pairwise Approach toClimate Data Homogenization
27 NOAA’s National Climatic Data Center
Step 5: Estimation of Step Change
• Use remaining metadata• Step-change magnitude calculated
according to model appropriate for each target-neighbor changepoint or as a simple difference in means
• Median of step estimates is used as adjustment; significance evaluated by estimating the 5th (median > 0) or 95th (median < 0) of pairwise estimates.
Briefing on Pairwise Approach toClimate Data Homogenization
28 NOAA’s National Climatic Data Center
σ (°C)
Simulated temperature series with random shifts caused by station moves/site changes
•Series in red treated as the target in subsequent figures
•All shifts are considered to be undocumented
•True “climate” trend in all simulated series is zero
(Annual Averages)
Briefing on Pairwise Approach toClimate Data Homogenization
29 NOAA’s National Climatic Data Center
Case 7 unadjustedTarget series and differences with neighbors before adjustment for undocumented shifts
Briefing on Pairwise Approach toClimate Data Homogenization
30 NOAA’s National Climatic Data Center
Target series and differences with neighbors after adjustment for undocumented shifts
Briefing on Pairwise Approach toClimate Data Homogenization
31 NOAA’s National Climatic Data Center
σ (°C)
Simulated temperature series following adjustment by pairwise algorithm
•Original Target Series in Red
•Adjusted Target Series in Green
•Adjusted Neighbor Series in Black
(Annual Averages)
Briefing on Pairwise Approach toClimate Data Homogenization
32 NOAA’s National Climatic Data Center
Diagnostic
• For the target example and its nine neighbors, 34 of 43 changepoints were detected and attributed to the correct series.
• Of the 9 changepoints not accounted for – 6 are under ±0.3σ– 2 are under ±0.5 σ– 1 was equal to 0.696σ (but was preceded by an
unidentified shift of -0.451 10 months earlier)
Briefing on Pairwise Approach toClimate Data Homogenization
33 NOAA’s National Climatic Data Center
Simulations
• Simulated 1000 groups of 21 correlated red noise series (n=1200)
• “Monthly Case 1”: Imposed between 0 and 10 changepoints at random locations and of random magnitude (average = 5)
• “Monthly Case 2”: As in case 1, except with random unrepresentative (“local”) trends (from 0.001σ/month up to about 0.18σ/month)
Briefing on Pairwise Approach toClimate Data Homogenization
34 NOAA’s National Climatic Data Center
Algorithm Results for “Step Change Only” Case
Briefing on Pairwise Approach toClimate Data Homogenization
35 NOAA’s National Climatic Data Center
Algorithm Results for “Steps and Local Trends” Case
Briefing on Pairwise Approach toClimate Data Homogenization
36 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
37 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
38 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
39 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
40 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
41 NOAA’s National Climatic Data Center
Impact of Adjustments on Trends
U.S. annual and seasonal temperature trends (°C dec-1) 1895 to 2006
0.0280.046-.0140.014S-O-N
0.0620.0640.0080.036J-J-A
0.0460.0590.0300.069M-A-M
0.0660.1010.0470.094D-J-F
0.0520.0700.0170.056Annual
UnadjustedAdjustedUnadjustedAdjusted
Minimum TemperatureMaximum TemperatureSeason
0.0280.046-.0140.014S-O-N
0.0620.0640.0080.036J-J-A
0.0460.0590.0300.069M-A-M
0.0660.1010.0470.094D-J-F
0.0520.0700.0170.056Annual
UnadjustedAdjustedUnadjustedAdjusted
Minimum TemperatureMaximum TemperatureSeason
Briefing on Pairwise Approach toClimate Data Homogenization
42 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
43 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
44 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
45 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
46 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
47 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
48 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
49 NOAA’s National Climatic Data Center
Briefing on Pairwise Approach toClimate Data Homogenization
50 NOAA’s National Climatic Data Center
How to conceive of the difference series?
function. step a is then , and in constant is if
)())(()(
vnT
YXYv
Xv
YvnT
XvnT
YvnT
XvnT
YXYv
XvvnT
D
nTD
Briefing on Pairwise Approach toClimate Data Homogenization
51 NOAA’s National Climatic Data Center
An Example: Reno, NevadaFrom: http://wattsupwiththat.wordpress.com/2008/01/10/how-not-to-measure-temperature-part-46-renos-ushcn-station/
Briefing on Pairwise Approach toClimate Data Homogenization
52 NOAA’s National Climatic Data Center
Reno, Nevada – Average Minimum Temperature
Briefing on Pairwise Approach toClimate Data Homogenization
53 NOAA’s National Climatic Data Center
Reno, Nevada – Average Minimum Temperature
Briefing on Pairwise Approach toClimate Data Homogenization
54 NOAA’s National Climatic Data Center
Reno, Nevada – Average Minimum Temperature
Briefing on Pairwise Approach toClimate Data Homogenization
55 NOAA’s National Climatic Data Center
Reno, Nevada – Average Minimum Temperature
Briefing on Pairwise Approach toClimate Data Homogenization
56 NOAA’s National Climatic Data Center
Reno, Nevada – Average Minimum Temperature
Briefing on Pairwise Approach toClimate Data Homogenization
57 NOAA’s National Climatic Data Center
Reno, Nevada
(a) Mean annual TOB and fully adjusted (TOB+Pairwise) minimum temperatures at Reno, Nevada(b) Difference between minimum temperatures at Reno and the mean from its 10 nearest neighbors.
Move toAirport
Likely urbanwarming
ASOS Equip.Moves
Briefing on Pairwise Approach toClimate Data Homogenization
58 NOAA’s National Climatic Data Center
Conclusions
• Pairwise comparison is the most direct way to detect undocumented changepoints
• Changepoint modeling is necessary in changepoint testing in order to identify unrepresentative trends
• No way to “safely” pass local (unrepresentative) trends through homogenization process
• Aliasing of trend inhomogeneities leds to a confused discussion about magnitude of UHI
Briefing on Pairwise Approach toClimate Data Homogenization
59 NOAA’s National Climatic Data Center
Future
• Adjust trend inhomogeneities as trends• Homogenize monthly data from Global
Historical Climatology Network• Derive daily adjustments for U.S. and
GSN/GHCN-Daily
Briefing on Pairwise Approach toClimate Data Homogenization