Base on NCEP/NCAR skin temperature (land surface temperature or skin temperature,LST) and ERA40 reanalysis surface soil temperature (ECMWF-STL1) data,the differences and similarities in spring and summer land surface thermal conditions are evaluated from the aspects of their spatial distribution and variability.Results show that:1) Temperature decreasing from the equator to the poles is found in the climatological distribution of spring land surface thermal conditions from both datasets.However,ECMWF-STL1 has higher temperatures than LST in mid-low latitude regions but lower in high latitude regions.In summer,the distributions of the multi-year mean of the thermal conditions of the global land surface that two data sets reflected are similar except for that over the Greenland.2) In spring,both of the inter-annual variability of ECMWF-STL1 and LST in high latitude of the Northern Hemisphere is larger than that in other regions.The inter-annual variability of ECMWF-STL1 is stronger than that of LST in northern Eurasia but shows an opposite pattern in the south part.In summer,significant inter-annual variability are found in central Africa,north Eurasia and some parts of the Americas,while there are evident differences in temperature variability in the North and South America areas between the two datasets.3) The first mode of EOF analysis suggests that a reversed phase of spring thermal condition variation between North and South Eurasia exists in both datasets.But the spatial patterns of thermal anomalies described by those two datasets are almost opposite in North America and Australia.In summer,the characteristics of two datasets are similar in Eurasian and Africa while not much the same in other regions.4) In spring,the significant warming happens in the high latitudes of Eurasia.Compared with LST,the warming trend of ECMWF-STL1 in northern Eurasia is more significant,while cooling trend in southern is relatively weak.In summer,the spatial pattern of the warming trend of ECMWF-STL1 is similar to that of LST,but with smaller values than LST.In short,the thermal conditions described by two datasets are almost consistent in most area of Africa and Eurasia,but show evident different in Australia,Greenland,America and especially over the Tibetan Plateau.