Using satellite remote sensing techniques to take quantitative observations of the climate system will advance our knowledge and ability to model the climate system and its changes. Polar-orbiting satellite records of global land surface skin temperature (LST) observations have high coverage and quality. Although limited in accuracy, and to a few years in length, these measurements are extremely useful because they are improvements on existing data.
This paper introduces a global, long-duration satellite diurnal cycle dataset of LST for 1981–98. LST measurements from National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) sensors are generally only twice per day for most areas and suffer problems such as cloud contamination and orbital drift. These problems are corrected using the algorithms developed by Jin and Dickinson, Jin, and Jin and Treadon, and consequently provide a long-term, monthly, global 8-km land surface skin temperature diurnal cycle dataset (LSTD) for temporally consistent LST records.
As a unique and independent resource, LSTD is valuable for studying the land surface climate and for evaluating model performance. For the first time, in this paper, the diurnal, seasonal, and interannual variations of LST are presented as a prototype use of this dataset. In addition, LSTD is compared with the National Centers for Environmental Prediction (NCEP)-National Center for Atmospheric Research (NCAR) reanalysis, in situ air temperature, the NOAA's Television Infrared Observation Satellite (TIROS) Operational Vertical Sounder (TOVS) observations, and shows encouraging consistencies on major features. These consistencies suggest that LSTD will be very valuable for climate change study.
Department of Meteorology, University of Maryland, College Park, College Park, Maryland