Using SCOOL Data

1
Using SCOOL Data
Dave Young CERES Co-Investigator and SCOOL Team
Member NASA Langley Research Center My NASA
Data Workshop August 2004
2
Topics

• SCOOL comparisons with CERES
• What de we compare?
• How well do the student observations compare with
  CERES?
• What does it mean?
• Using SCOOL Data
• How do we access the data?
• What types of studies can we perform with SCOOL
  data?

3
Clouds and the Earths Radiant Energy System
How is the global energy budget changing? Do
human-made aerosols affect the Earths
climate? Do clouds amplify or reduce climate
change?
4
Energy Balance of the Earth
5
CERES/SCOOL ComparisonsWhat is Compared?

• Cloud amount
• Clear (0-5)
• Partly Cloudy (5-50)
• Mostly Cloudy (50-95)
• Overcast (95-100)
• Cloud height
• Low (below 2000 meters)
• Mid (between 2000 and 6000 meters)
• High (above 6000 meters)

• Cloud layers
• Cloud Visual Opacity (how much light the cloud
  allows through)

6
Terra Comparison ExampleGloucester, VA June 4,
2001(http//scool.larc.nasa.gov/query_data.html)
7
GOES Visible Image June 4, 2001 1645 GMT
8
Second Terra Comparison ExampleGloucester, VA
June 4, 2001
9
GOES Visible Image June 7, 2001 1715 GMT
10
Clear Day ComparisonOctober 20, 1997 1745 GMT
Visible
Infrared
11
Comparison ResultsOctober 20, 1997 1745 GMT
Comment Comparisons of Clear-Sky Observations
Are Very Important!
12
Cloudy Day Comparison October 22, 1997 1745 GMT
Visible
Infrared
13
Comparison ResultsOctober 22, 1997 1745 GMT
Comment Poquoson observed contrails - These may
be what the Newport News students were reporting.
There also was a mid-level layer cloud to the
south.
14
Status of SCOOL Database (as of July 2003)

• Over 32,000 student observations
• Matching with CERES
• Matched using reported UTC
• Observations must be within 15 minutes
• CERES data from TRMM, Terra, and Aqua
• A total of 5409 matches
• Why only 5409 out of 32,000? (only 17)
• Class times not in sync with satellites?
• Time reporting errors?
• All CERES data not yet processed

15
CERES TRMM/SCOOL Cloud Amount Comparison(Observa
tions within 15 minutes)
16
Summary of Cloud Comparison

• The Students and Satellite Observations
• Agree 60 out of 99 times (60.6)
• Weakly Agree 29 out of 99 times (29.3)
• Disagree 9 out of 99 times (9.1)
• Strongly Disagree 1 out of 99 times (1.0)

17
CERES Terra/SCOOL Cloud Amount
Comparison(Observations within 15 minutes)
18
CERES Terra/SCOOL Cloud Layer Comparison
19
Major results

• Overall agreement is good (but not great)
• Many mismatched points are caused by time errors
• 3-class errors are frequently due to spatial
  mismatch (7 of 22 3-class errors from one site
  near edge of grid)
• No significant increase in cloud cover errors
  over snow
• Major discrepancy in layers due to thin high
  clouds over low

20
Accessing and Using SCOOL Data
21
Downloading SCOOL Data

• SCOOL Web Site
• http//asd-www.larc.nasa.gov/SCOOL/
• Accessing Data
• http//scool.larc.nasa.gov/en_query_data.html

22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
SCOOL / CERES Comparison
28
Example

• Sample Science Questions
• How Often Do Students See More Than One Layer of
  Clouds?
• Do Low Clouds Cover More of the Sky Than High
  Clouds?
• Method
• Download data from Virginia from April 2001
• Count how often each layer occurs
• automate in Excel or count manually
• Plot

29
(No Transcript)
30
(No Transcript)
31
Low Cloud Occurrence
32
How Often Do Students See More Than One Layer of
Clouds?
33
Do Low Clouds Cover More of the Sky Than High
Clouds?
Low Clouds
High Clouds
34
Do Low Clouds Cover More of the Sky Than High
Clouds?
35
Other SCOOL Data Studies

• Compare cloud properties from different seasons
• Do we expect to see differences?
• What is the reason for the differences?
• Compare cloud properties from different schools
• Schools can be from different countries or nearby
• Can there be significant cloud differences from
  nearby schools?
• Compare morning and afternoon cloud properties
• The SCOOL Team is VERY interested in the results!

36
Extras
37
Understanding the Comparisons

• The Importance of Time Matching
• The Need for Many Different Observations
• Guess the Cloud Amount Activity
• The Comparison with Random Observations

38
(No Transcript)
39
(No Transcript)
40
Results From 50 Dice Rolls
41
Results From 500,000 Dice Rolls
42
Statistical Comparison

• Chi-squared 8.3
• Probability of Occurring by Chance 49.9

• Chi-squared 77.4
• Probability of Occurring by Chance
  0.00000000005