Analysis of NearSurface Temperature

1
Analysis of Near-Surface Temperature in Areas of
Known Land Use/Land Cover Change Rob Hale CIRA,
Colorado State University Kevin Gallo Center for
Satellite Applications and Research, NOAA/NESDIS
Tim Owen NOAA National Climatic Data Center
2
Overview

• Purpose to utilize a unique source of land
  use/land cover (LULC) data in analyzing the
  effects of LULC change on near-surface
  temperature measurements
• Data and Analysis
• USGS Land Cover Trends
• U.S. Climate Normals stations
• NCEP-NCAR 50-Year Reanalysis
• Results
• Minimum Temperatures
• Maximum Temperatures
• Conclusions

3
USGS Land Cover Trends Project

• 10km or 20km square sample blocks randomly
  selected within ecoregions for LULC analysis
• Satellite and aerial imagery and ancillary data
  used for supervised classification into 11 LULC
  types
• Nominal years of analysis 1973, 1980, 1986,
  1992, and 2000
• 31 ecoregions completed thus far (857 sample
  blocks)

4
An Example Mississippi Valley Loess Plains
Ecoregion
1972 MSS
Memphis WSFO
Crop/Pasture ? Urban Forest ? Urban Water ?
Urban
2000 TM
5
U.S. Climate Normals Stations

• 5332 Stations
• Monthly maximum and minimum near-surface
  temperature observations
• Corrected for time of observation biases and
  station and instrument changes
• Analyzed anomalies from the 1971-2000 means

6
U.S. Climate Normals Stations

• Identified stations (526) within 10km of Land
  Cover Trends sample blocks
• Determined the period of greatest single type of
  LULC change and what that conversion was

7
U.S. Climate Normals Stations

• Identified stations (526) within 10km of Land
  Cover Trends sample blocks
• Determined the period of greatest single type of
  LULC change and what that conversion was
• Limited analyses to stations with at least 2
  years of data (both Normals and NNR) before and
  after period of greatest LULC change (130)

8
U.S. Climate Normals Stations

• Identified stations (526) within 10km of Land
  Cover Trends sample blocks
• Determined the period of greatest single type of
  LULC change and what that conversion was
• Limited analyses to stations with at least 2
  years of data (both Normals and NNR) before and
  after period of greatest LULC change (130)
• Stations also limited to elevations 500m (94)

9
NCEP-NCAR 50-Year Reanalysis (NNR) Data

• Gridded estimates of monthly minimum and maximum
  near-surface temperature at 209km resolution
• Estimates based on satellite and rawinsonde upper
  air data no input of surface-based temperature,
  moisture, or wind observations
• Values of nearest gridpoints bilinearly
  interpolated to Normals station locations
• Anomalies from 1979-2000 means utilized for
  comparison

10
Results Normals Minimum Temperatures

• Post-change minimum temperatures warmer for all
  change types, significant difference for many
  types
• Overall, significantly warmer post-change minima

11
Results NNR Minimum Temperatures

• As with Normals minimum temperatures, many LULC
  conversion types associated with significantly
  warmer post-change period minima
• Post-change periods significantly warmer overall

12
Results Normals-NNR Differences in Minimum
Temperatures

• Two LULC conversion types (both conversions to
  urban) show significantly more post-change
  warming in the Normals observations than in the
  NNR data
• Overall, Normals minimum temperatures showed
  significantly greater increases following LULC
  conversion than did NNR data for the same periods

13
Results Normals-NNR Differences in Minimum
Temperatures

• Some LULC conversions that might be expected to
  produce significant differences did not

14
Results Normals Maximum Temperatures

• Warmer maximum temperatures following LULC change
  prevail, but not as ubiquitous as for minimum
  temperatures
• For all stations combined, maximum temperature
  significantly warmer for post-change period

15
Results NNR Maximum Temperatures

• Post-change periods mostly warmer than pre-change
  periods many significantly warmer
• Warming again not as ubiquitous as for minimum
  temperatures

16
Results Normals-NNR Differences in Maximum
Temperatures

• Results more mixed than for minimum temperatures
  overall difference not significant
• Significantly more post-change warming for
  Normals than for NNR for two conversion types
  (although n 2 for mechanically disturbed ?
  forest)

17
Results Normals-NNR Differences in Maximum
Temperatures

• Again, some maybe unexpected results, including
  forest ? urban
• Consistent with prior research showing the UHI
  effect is more pronounced for minimum
  temperatures than maximum temperatures

18
Conclusions

• Urbanization of crop/pastureland resulted in
  significant warming of both minimum and maximum
  temperatures
• Urbanization of forested land resulted in
  significant warming of minimum temperatures (but
  insignificant warming of maximums)
• Clear-cutting of forests and afforestation of
  crop/pastureland did not result in significant
  changes (min or max)

19
Conclusions

• As more ecoregions are completed by the Land
  Cover Trends Project, more stations can be
  analyzed to provide greater statistical
  confidence in the results