GC21A-1076
Comparison of MODIS Satellite Land Surface Temperature with Air Temperature along a 5000-metre Elevation Transect on Kilimanjaro, Tanzania.

Tuesday, 15 December 2015
Poster Hall (Moscone South)
Nicholas C Pepin, University of Portsmouth, Portsmouth, PO1, United Kingdom, Ryan Williams, University of Portsmouth, Portsmouth, United Kingdom and Eduardo Eiji Maeda, University of Helsinki, Helsinki, Finland
Abstract:
There is concern that high elevations may be warming more rapidly than lower elevations, but there is a lack of observational data from weather stations in the high mountains. One alternative data source is satellite LST (Land Surface Temperature) which has extensive spatial coverage. This study compares instantaneous values of LST (1030 and 2230 local solar time) as measured by the MODIS MOD11A2 product at 1 km resolution with equivalent screen level air temperatures (in the same pixel) measured from a transect of 22 in situ weather stations across Kilimanjaro ranging in elevation from 990 to 5803 m. Data consists of 11 years on the SW slope and 3 years on the NE slope, equating to >500 and ~140 octtads (8-day periods) respectively. Results show substantial differences between LST and local air temperature, sometimes up to 20C. During the day the LST tends to be higher than air temperature and the reverse is true at night. The differences show large variance, particularly during the daytime, and tend to increase with elevation, particularly on the NE slope of the mountain which faces the sun when the daytime observations are taken (1030 LST). Differences between LST and air temperature are larger in the dry seasons (JF and JJAS), and reduce when conditions are more cloudy. Systematic relationships with cloud cover and vegetation characteristics (as measured by NDVI and MAIAC for the same pixel) are displayed. More vegetation reduces daytime surface heating above the air temperature, but this relationship weakens with elevation. Nighttime differences are more stable and show no relationship with vegetation indices. Therefore the predictability of the LST/air temperature differences reduces at high elevations and it is therefore much more challenging to use satellite data at high elevations to complement in situ air temperature measurements for climate change assessments, especially for daytime maximum temperatures.