Global hourly land surface air temperature datasets: inter‐comparison and climate change

• Published in: International Journal of Climatology
• Main Authors: Wang, Aihui, Zeng, Xubin
• Other Authors: National Natural Science Foundation of China, National Science Foundation, U.S. Department of Energy, National Aeronautics and Space Administration
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2015
• Subjects: Antarc*; Antarctica; Greenland
• Online Access: http://dx.doi.org/10.1002/joc.4257; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1002%2Fjoc.4257; https://onlinelibrary.wiley.com/doi/pdf/10.1002/joc.4257; https://onlinelibrary.wiley.com/doi/full-xml/10.1002/joc.4257; https://rmets.onlinelibrary.wiley.com/doi/am-pdf/10.1002/joc.4257; https://rmets.onlinelibrary.wiley.com/doi/pdf/10.1002/joc.4257

ABSTRACT The land surface air temperature ( LSAT ) is one of the fundamental parameters to represent heat transfer and to modulate the moisture cycle between land and atmosphere. Here, we quantify the spatiotemporal characteristics of four newly developed hourly temperature products by merging reanalysis with in situ data Climatic Research Unit ( CRU ). Overall, LSATs from different hourly products are consistent with each other, and their differences are generally smaller in magnitude than biases between hourly products and monthly averaged daily maximum and minimum temperature data from CRU . While the true monthly mean (using hourly values) and the monthly mean (of daily maximum and minimum temperatures) and their seasonality [as represented by the (July–January) differences] differ, their trends agree with each other very well. The polar amplification ratio of average temperature trend north of 65°N to that over global land (excluding Greenland and Antarctica) is also similar among different products, with the annual ratio of around 1.7. The ratio in summer (June–August) is always smaller than the annual value for different periods among all products. Based on the probability distribution functions from the monthly anomalies of different variables, the coldest tenth percentile of temperature in each decade overall increases with time, while the warmest tenth percentile does not vary much from 1950–1979, followed by a rapid increase from 1980–2009. These results and additional sensitivity tests suggest that the 4‐h LSAT products can be widely used for climate analysis, model evaluation, and offline land surface modelling from 1948–2009.