magicjava

Aqua Satellite Data Processing

A Note On The UAH And RSS Raw Data

How UAH And RSS Temperatures Are Measured

Overview Of The Aqua Satellite

Looking At The Aqua Satellite Data

UAH Satellite Data

Dangit! More Climate Stuff. UAH and RSS Raw Data

Compute Radiance From Frequency And Temperature
This code converts a reading in degrees Kelvin and a frequency value to a radiance value. Radiance is given in units of milliwatts/m² cm^-1. The original C code can be found in the file read_chan_props.c in the C Readers Code provided by NASA. The example here uses pseudocode.

Function ComputeRadiance

   Parameter Temperature Of Type Double

   Parameter Frequency Of Type Double

   Return Value Radiance Of Type Double

Begin Function

   Constant PLANCKS_CONSTANT Equals 6.62620e-34

   Constant SPEED_OF_LIGHT Equals 2.99793e+10-34

   Constant BOLTZMANS_CONSTANT Equals 1.38062e-23

   Constant PLANK_CONSTANT_1 Equals 2.0 * (PLANCKS_CONSTANT) * (SPEED_OF_LIGHT) * (SPEED_OF_LIGHT) * 1.e7

   Constant PLANK_CONSTANT_2 Equals (PLANCKS_CONSTANT) * (SPEED_OF_LIGHT) /

(BOLTZMANS_CONSTANT)



   Return PLANK_CONSTANT_1 * Frequency * Frequency * Frequency / (Exponent(PLANK_CONSTANT_2 * Frequency / Temperature) - 1.0)

End Function

Adjusting Footprint Brightness
Recall that the AMSU scans 30 footprint readings per scan line. Because theses readings are done with the satellite in the same position, footprints to the far left and far right will be at different heights in the atmosphere than readings directly below the satellite. A visual example of these offsets is shown in the diagrams below. Diagram A shows the offsets for channels 1-3, and 15. Diagram B shows the offsets for the other AMSU channels.

These offsets cause a deviation in the brightness values for each footprint. These deviations are shown in the diagram below. The X axis shows the footprint readings. The Y axis shows the variations in brightness that occur at each footprint for each channel.

The values in this chart can be used to adjust the brightness readings of each footprint in a scan line of the AMSU.

These diagrams, as well as a more complicated way of adjusting brightness values, are provided in The Limb Adjustment of AMSU-A Observations: Methodology and Validation.

Pseudocode for using the brightness adjustment diagram provided here is shown below.

Foreach Channel

   Foreach Footprint

      Subtract Channel/Footprint Adjustment Value From Channel Brightness Reading

   End Foreach

End Foreach

NOTE: These adjustments should be done to the Brightness values, not the antenna values.

References:

Summary Of Aqua Satellite Data, Computer Code, And Broken Equipment

Aqua Satellite Raw AMSR-E Data

Aqua Satellite Data Processing

A Note On The UAH And RSS Raw Data

How UAH And RSS Temperatures Are Measured

Overview Of The Aqua Satellite

Looking At The Aqua Satellite Data

UAH Satellite Data

Dangit! More Climate Stuff. UAH and RSS Raw Data

C AMSU Level 1B Swath and Grid Readers
The Limb Adjustment of AMSU-A Observations: Methodology and Validation

Tuesday, February 2, 2010

More On Water Vapor, Length Of Day Delta Correlation

I downloaded The official Length of Day Delta (LODD) data so that I could do a more detailed analysis on the correlation between water vapor and the LODD I discussed here. I parsed up the data and added it to my working copy of the Climate Scientist Starter Kit Spreadsheet.

What I found is the correlation is not as good as I first thought it was. Or maybe a better way to say it is the correlation is different than I first thought it was. There seems to be two correlations in the data.

Just to be clear, I'm not suggesting that changes in the LODD are causing changes in water vapor. I'm pointing out that there are correlations between the two data sets.

What Is Length Of Day Delta (LODD)?
Let's start with an explanation of what the Length Of Day Delta (LODD) is. Simply put, the LODD is a measure of how much extra time is added to each day. A "normal" day has 86,400 seconds. However, in practice, there's no such thing as a normal day. Each day has a few more or a few less milliseconds. The extra amount of time, which can be positive or negative, is the LODD.

Water Vapor And LODD, 1983 To 2008

The graph above shows water vapor (blue lines) and LODD (green lines) data for the period of July, 1983 to June, 2008. There's a 13 month running mean added to smooth the data.

As you can see, the correlation doesn't look as nice as when I just overlaid two graphs in my previous post. Specifically, in the early years water vapor is not matching LODD. I was a bit bummed out when I saw this, thinking a correlation that looked so promising turned out to be nothing.

But I kept looking and found something else: two correlations.

Water Vapor And LODD, 1983 To 1997

The first correlation runs through the first half of the data, from July 1983 to December, 1997. In this correlation the water vapor (blue lines) and LODD (green lines) are mirror images of each other. As one goes up, the other goes down. This is true both at the monthly level and at the 13 month smoothed trend.

Water Vapor And LODD, 1998 To 2008

The second correlation runs through the second half of the data, from January, 1998 to June, 2008. In this correlation water vapor (blue lines) and LODD (green lines) follow each other in the 13 month smoothed trend, even though at the monthly level they are still have reversed trends.

What's It All Mean?
The honest answer is "I don't know". Maybe there's something to this, maybe not. But the two correlations seem pretty clear.

One thing that jumps out at me is that 1998, when the new smoothing trend starts, is when the Earth stopped warming.

I'll continue to research this to see what I find and post updates here.

References:
Earth Orientation Centre Length of Day Delta (LODD) data
Stunning New Climate Correlation
Climate Scientist Starter Kit Spreadsheet

Monday, February 1, 2010

Stunning New Climate Correlation

I just discovered this and simply had to post it. It's an amazing correlation between water vapor in the Earth's atmosphere and changes in the Earth's rotation measured in milliseconds! I know. I know. It's insane to think changes of a few milliseconds could somehow be related to the total water vapor in the air. But here's the chart:

The blue line is changes in the Earth's rotation as measured in milli-seconds. The red line is changes in the total amount of water vapor in the Earth's atmosphere. The graph covers a 24 year period from 1983 through 2007.

The information for changes in the Earth's rotation comes from the U.S. Naval Observatory. The information for changes in the Earth's water vapor comes from the International Satellite Cloud Climate Project. You can graph the cloud data using the Climate Scientist Starter Kit and your favorite spreadsheet program.

I don't know what this means yet. As I said, I only just now discovered it. But I wanted to get the information out there so people can see it.

Rest assured that future posts will look into this more deeply.

References:
U.S. Naval Observatory Leap Seconds
International Satellite Cloud Climate Project
Climate Scientist Starter Kit

Summary Of Aqua Satellite Data, Computer Code, And Broken Equipment

Availability Of Satellite Data And Associated Computer Code
The following tables sum up the availability of data and computer code used to produce the UAH, RSS, and AMSR-E information we've been discussing in this series of articles.

UAH Data	Available To Public
Data Transmitted From Satellite To Ground	No
Level 0 Data	Partial¹
Level 1A Data	No
Level 1B Data	Yes
Level 2A Data	Yes

UAH Computer Code	Available To Public²
Process Satellite Data To Level 0	No
Process Level 0 To Level 1A	No
Process Level 1A To Level 1B	No
Process Level 1B To Level 2A	No

RSS Data	Available To Public
Data Transmitted From Satellite To Ground	No
Level 0 Data	No
Level 1A Data	No
Level 1B Data	No
Level 2A Data	No

RSS Computer Code	Available To Public²
Process Satellite Data To Level 0	No
Process Level 0 To Level 1A	No
Process Level 1A To Level 1B	No
Process Level 1B To Level 2A	No

AMSR-E Data	Available To Public
Data Transmitted From Satellite To Ground	No
Level 0 Data	Partial¹
Level 1A Data	Yes
Level 1B Data	No
Level 2A Data	Yes

AMSR-E Computer Code	Available To Public²
Process Satellite Data To Level 0	No
Process Level 0 To Level 1A	No
Process Level 1A To Level 1B	No
Process Level 1B To Level 2A	Yes

Notes
1 UAH and AMSR-E Level 0 data is available for a 4 day window from the University of Wisconsin. However, there seems to be no read permissions allowed to guest accounts for files other than 00README.txt.
2 Computer source code to read and write various versions of HDF-EOS files is publicly available. However, this code does not actually process climate data.

Broken Satellite Equipment
The following tables sum up the satellite equipment used to produce the UAH, RSS, and AMSR-E information that has broken over time.

Broken UAH Equipment

Channel 7

Broken RSS Equipment

Unknown

Broken AMSR-E Equipment

Hot Load

Horn A

References And Additional Posts In This Series

Aqua Satellite Raw AMSR-E Data

Aqua Satellite Data Processing

A Note On The UAH And RSS Raw Data

How UAH And RSS Temperatures Are Measured

Overview Of The Aqua Satellite

Looking At The Aqua Satellite Data

UAH Satellite Data

Dangit! More Climate Stuff. UAH and RSS Raw Data

Aqua Satellite Raw AMSR-E Data

In this post we take a look at the AMSR-E Level 2A data. Specifically, we now want to be able to write pseudocode that reads the file, checks for errors, and grabs the raw temperature data. In order to get to this level of understanding of the data, we're going to use the Level 2A Data Fields web page. It describes the fields in the Level 2A data file for the AMSR-E instrument.

Additional Posts In This Series

Aqua Satellite Data Processing
A Note On The UAH And RSS Raw Data
How UAH And RSS Temperatures Are Measured
Overview Of The Aqua Satellite
Looking At The Aqua Satellite Data
UAH Satellite Data
Dangit! More Climate Stuff. UAH and RSS Raw Data

Scan Information

Each AMSR-E Level 2A data file contains a singe scan at low resolution and a single scan at high resolution.

For low resolution, the AMSR-E scans 1994 times in the same direction as the satellite is moving. Each of these 1994 scans contains 243 readings perpendicular to the movement of the satellite (called Footprints), and each of these 243 readings contains 5 channels, each at a different frequency.

This gives a total of 2,422,710 readings in a single Level 2A file. This data has been "smoothed" as part of its processing from Level 1B to Level 2A.

For high resolution scans the AMSR-E scans 1994 times in the same direction as the satellite is moving. Each of these 1994 scans contains 486 readings perpendicular to the movement of the satellite, and each of these 486 readings contains 1 channel at a frequency of 89.0 GHz.

This gives a total of 969,084 readings in a single Level 2A file. This data has not been smoothed. It's in its original Level 1B form.

NOTE: The high resolution scans were designed to be collected by two "horns" on the AMSR-E instrument, called Horn A and Horn B. However, recall from this post that Horn A failed on November 11th, 2004. A software change was made to gather all high resolution data from Horn B. Horn A still has data fields in the level 2A file, but they are all set to -32,768.

NOTE: Both the low resolution and high resolution temperatures are stored in a scaled offset format. To obtain the actual temperature in degrees Kelvin multiply the stored values by 0.01 and add 327.68 to the result.

Data Quality

There are several fields in the AMSR-E Level 2A data file that contain the results of Quality Assurance (QA) checks. These fields tell if various pieces of data in the file can be used with confidence. Here are the important QA fields to check when processing temperature data:

For Low Resolution Scans

Scan Quality Flag A series of 1994 flags indicating the quality of each scan line. If a flag is not zero, the corresponding scan line should be rejected.
Channel Quality Flag 6 To 52 A series of 1994 arrays containing 12 flags each. There are two flags for channels 6.9, 10.7, 18.7, 23.8, and 36.5. The last two fields are not used as there is no channel 52. These flags indicate the horizontal and vertical scan validity for each scan footprint. If either flag for a footprint is not zero, the corresponding footprint should be rejected.

For High Resolution Scans

There are two high resolutions scans in the Level 2A file, one for Horn A and one for Horn B. Horn A is broken and its data should be ignored.

Scan Quality Flag 89B A series of 1994 flags indicating the quality of each scan line for Horn B. If a flag is not zero, the corresponding scan line should be rejected.
Channel Quality Flag 89B A series of 1994 arrays containing 2 flags each. These flags indicate the horizontal and vertical scan validity for each scan footprint. If either flag for a footprint is not zero, the corresponding footprint should be rejected.

Pseudocode For Reading Raw AMSR-E Data

The following pseudocode shows how to process a AMSR-E Level 2A file, checking for QA errors, and grabbing the temperature data. Recall that temperatures are provided in scaled offset degrees Kelvin and that Horn A is always bad.

Open File

Read Low Resolution Record Into Memory



Foreach Of The 1994 Scanlines

   If Scan Quality Flag Is Not 0 Then Reject Scanline



   Foreach Of The 5 Footprints

      If Either Of The Channel Quality Flags Are Not 0 Then Reject Footprint



      Read Each Of The 243 Footprints For the Channel

   End Foreach

End Foreach



Read High Resolution Horn B Record Into Memory



Foreach Of The 1994 Scanlines

   If Scan Quality Flag Is Not 0 Then Reject Scanline



   If Either Of The Channel Quality Flags Are Not 0 Then Reject Footprint



   Read Each Of The 486 Footprints For the Channel

End Foreach

References: