5. TRMM - Results
In order to understand the impacts of the cleared areas surrounded by forest in the region of study, the rainfall rate of the TRMM satellite was used. The rainfall rate was estimated from the combined TMI and PR algorithms shown below.
The steps used in the algorithms
are described in the table below:
| The TMI calibration algorithm (1B-11) converts the radiometer counts to antenna temperatures by applying a linear relationship of the form Ta = c1 + c2 x count. The coefficients are provided by the instrument contractor. Antenna temperatures are corrected for cross-polarization and spill over to produce brightness temperatures (Tb), but no antenna beam pattern correction or sample to pixel averaging are performed. Temperatures are provided at 104 scan positions for the low frequency channels and 208 scan positions at 85 GHz. There are four samples per pixel (3 -dB beamwidth) at 10 GHz, two samples at 19, 22, and 37 GHz, and one sample per pixel for the 85 GHz. | The PR calibration algorithm (1B-21) converts the counts of radar echoes and noise levels into engineering values (power) and outputs the radar echo power and noise power separately. The algorithm also detects and flags the range bin with return power that exceeds a pre-determined threshold value. |
| The PR reflectivity algorithm (1C-21) converts the power and noise estimates from 1B-21 to radar reflectivity factors (Z-factors). In order to reduce output data volume, only pixels with power that exceeds the minimum echo detected in 1B-21 are converted and stored. | The PR qualitative algorithm (2A-23) inputs PR reflectivities (1C-21) and returns a rain/no rain decision based on echo structure. When rain is present, a storm height is calculated. If a bright band is detected, the height of the bright band is also given. |
| The TRMM combined algorithm (2B-31) combines data from the TMI and PR to produce the best rain estimate for TRMM. Currently, it uses the low frequency channels of TMI to find the total path attenuation. This information is used to constrain the radar equation. Inputs for this algorithm include TMI radiance (1B-11), PR reflectivity (1C-21), and PR qualitative (2A-23). | |
The figure below is a Landsat satellite
image from the Rondonia state in Brazil. This shows the forest-cleared
areas of this study. The black horizontal line indicates the cross-section
where the rainfall rate was evaluated.
LANDSAT satellite image over Rondonia,
Brazil (Calvert et al., 1997).
Channels: TM1(520-600nm), TM2(520-600nm),
TM3(630-690nm), TM4(760-900nm), TM5(1550-1750nm), TM7(2080-2350nm)
The figure below shows a west-east
cross-section of the rainfall rate estimated by the TRMM satellite.
Intriguing maximuns are found over
the forest areas near the borders of forest-cleared regions.
It suggests a possible relationship
with the surface changes over the area in agreement with our
hypothesis posted in the introduction.
However, a more careful study has to be done in order
to confirm these results.
West-east cross-section of TRMM rainfall rate (mm/hr) averaged for January and February 1999.