An Evaluation and Comparison of Vertical Profile Data from the VISSR Atmospheric Sounder (VAS)

View More View Less
  • 1 Universities Space Research Association, Atmospheric Sciences Division, Marshall Space Flight Center, Huntsville, AL 35812
© Get Permissions
Full access

Abstract

Statistical procedures are used to compare vertical profiles of temperature and moisture derived from VAS with three different algorithms to those of corresponding rawinsonde measurements for a clear-cold environment. To account for time and space discrepancies between the data, rawinsonde values were adjusted to the satellite sounding times. Both rawinsonde and satellite soundings were objectively analyzed onto a mesoscale grid. These grid point values were compared at 50 mb pressure increments from the surface up to 100 mb. The data were analyzed for horizontal and vertical structure, representativeness of derived parameter, and significant departure (improvement) from the a priori (first guess) information.

Results indicate strong temperature and moisture biases in the satellite soundings. Temperature biases of 1–4°C and dew-point biases of 2–6°C generally occur in layers where strong inversions are present. Magnitudes vary with time as the atmospheric features evolve. The biases change as a function retrieval scheme, suggesting limitations and restrictions on the applications of the various techniques. Standard deviations of temperature range from 1–2°C for each retrieval scheme with maxima near 800 and 400 mb. Derived parameters (precipitable water and thickness) suffer from similar biases, though to a somewhat lesser extent. Although satellite-derived gradients of basic and derived parameters are generally weaker than those from rawinsondes, they have good horizontal structure where magnitudes of the parameters are relatively strong. Integrated thermal (thickness) and moisture (precipitable water) parameters show mixed results. Although biases are small in the precipitable water values from the regression scheme, horizontal structure is poor.

Analysis of first guess information shows similar biases when compared to the ground truth measurements. This information, however, seems to provide the majority of the vertical structure present in the VAS retrievals.

Abstract

Statistical procedures are used to compare vertical profiles of temperature and moisture derived from VAS with three different algorithms to those of corresponding rawinsonde measurements for a clear-cold environment. To account for time and space discrepancies between the data, rawinsonde values were adjusted to the satellite sounding times. Both rawinsonde and satellite soundings were objectively analyzed onto a mesoscale grid. These grid point values were compared at 50 mb pressure increments from the surface up to 100 mb. The data were analyzed for horizontal and vertical structure, representativeness of derived parameter, and significant departure (improvement) from the a priori (first guess) information.

Results indicate strong temperature and moisture biases in the satellite soundings. Temperature biases of 1–4°C and dew-point biases of 2–6°C generally occur in layers where strong inversions are present. Magnitudes vary with time as the atmospheric features evolve. The biases change as a function retrieval scheme, suggesting limitations and restrictions on the applications of the various techniques. Standard deviations of temperature range from 1–2°C for each retrieval scheme with maxima near 800 and 400 mb. Derived parameters (precipitable water and thickness) suffer from similar biases, though to a somewhat lesser extent. Although satellite-derived gradients of basic and derived parameters are generally weaker than those from rawinsondes, they have good horizontal structure where magnitudes of the parameters are relatively strong. Integrated thermal (thickness) and moisture (precipitable water) parameters show mixed results. Although biases are small in the precipitable water values from the regression scheme, horizontal structure is poor.

Analysis of first guess information shows similar biases when compared to the ground truth measurements. This information, however, seems to provide the majority of the vertical structure present in the VAS retrievals.

Save