Products - Satellite vs Model

1. Application of Forward Model on Satellite Observation Bias Identification

DMSP F-16 SSMIS Ch. 4 (54.4 GHz) brightness temperature bias is observed by directly comparing satellite observations to forward model simulation using NCEP GFS results as background fields. It is clearly shown that positive anomalies exist over middle to high latitude in North Hemisphere in ascending node and middle to low latitudes of both hemispheres in descending node (Fig. 1). It is also found that similar anomalies happen in all low atmosphere sounder (LAS) channels. Further investigations discover that both antenna reflector emission and solar intrusion on warm calibration targets contribute to such bias. Cal/Val group in NOAA/NESDIS develops a set of correction algorithm to identify the error and successfully remove it so that the data can be used for NWP operational missions, environment data retrieval, and climate change studies.

2. Application of Forward Model on Cross-Calibration for Climate Change Study

The identification of inter-sensor bias is an major issue in order to use different satellite observations on climate change study. A few identification methods have been developed to try resolving this challenge. However, those methods either provide small number of data samples or are over limited regions. To make more samples over global area, double difference technique (DDT) is brought up. DDT uses difference of observation and model biases from two individual satellites to represent inter-sensor difference. In this method, forward model plays a key role in inter-sensor bias detection when some restrictions are applied. Preliminary tests in DMSP SSM/I channels prove that DDT can be successfully used for microwave channel cross-calibration for climate change study.