This paper presents a comprehensive analysis of the patterns of spatio-temporal variations in the Mars Surface Brightness Temperature（SBT） by using thermal infrared data acquired by three orbiters. Inter-annual changes in SBT are determined by comparing temperature data from different thermal infrared sensors

with spectral and local time corrections. Seasonal variations in the northern（NH） and southern（SH） hemisphere night SBT are presented with four Mars-year Thermal Emission Spectrometer（TES） data. The influences of the latitude and altitude on night SBT are analyzed based on TES measurements. The discrepancies between the adjusted Infrared Thermal Mapper（IRTM） band-B and the Mars Climate Sounder（MCS） band-A4 measurements are 1.3 K and 1.0 K for day and night SBT

respectively.The differences between the IRTM band-B and the average TES Mars Year（MY） 24—26 are 3.1 K and 2.1 K for day and night SBT

respectively. During the aphelion period

seasonal changes in SH night SBT conform well to the sine curve

and the inter-annual variation is lower than 3 K. However

the seasonal variation greatly diverges from the sine curve; the inter-annual difference is significant during the perihelion period. By contrast

NH night SBT exhibits no clear seasonal trend; the variation of this parameter is smaller than that of SH. No signs of the sine-curve pattern are displayed. Night SBT increased by approximately 17 K in NH and SH during the MY 25 global dust storm. When Ls is 115°—125°

NH SBT is obviously influenced by topography

whereas the SH SBT isotherms are nearly parallel to the latitude lines. Some plains

such as Acidalia Planitia

Chryse Planitia

Isidis Planitia

and Utopia Planitia

are warmer than their surrounding areas; moreover

high-altitude regions

including Arsia Mons

Alba Patera

Elysium Mons

Terra Sabaea

and Olympus Mons

have almost the lowest SBT

except for the Polar regions. When Ls is 295°—315°

the SH isotherms are fragmentized

which corresponds to the cratered surfaces. The NH isotherms in winter are only parallel to the latitude lines north of 50° N. The maximum SBT gradually decreased from the equator to the poles

except for 35° S. The minimum SBT between 15° S and 35° N are lower than that of the higher latitude regions because of the influence of to pography. The SBT standard deviations and differences between maxima and minima reveal the same law

that is

the closer to the equator

the higher are both values. Conclusions: The following conclusions are reached.（1） The 5 Mars-year inter-annual variation in SBT is within the precision range of measurement and processing; thus

the detection of Mars climate change is not feasible using current thermal infrared measurements.（2） The inter-annual difference in SBT during the perihelion period is higher than that during the aphelion period. The amplitude of seasonal variations in NH is smaller than that in SH. Increased SBT in NH caused by dust storm is higher than that caused by increased solar radiation in summer. Low-altitude topographies

such as basin sand canyons

exhibit higher temperatures than high-altitude topographies

such as mountains

terrains

and plateaus. Night SBT exhibits a strong negative correlation with altitude in Olympus Mons. The linear fitting result indicates that night SBT decreases by approximately 1.4 K as the altitude increases by 1km. Night SBT at low latitudes is generally higher than that at high latitudes. Influenced by factors

such as topography

the SBT maximum exists at 35° S instead of the equator. The SBT minima at the latitudes close to the equator are smaller than those at high latitudes. The lowlatitude SBT varies more intensely than those near the polar latitudes.