The effect of land use land cover types on MODIS land surface temperature in Ghana

Abstract

In this study we examine the effect of land use land cover (LULC) types on MODIS land surface temperature (LST) in Ghana. The purpose of the research is to explore the relationship between air temperature ([Formula: see text]) and LST over LULC types and to identify their association with elevations. Datasets consist of MODIS land cover type-product (MCD12Q1), Shuttle Radar Topography Mission (SRTM) DEM, and composite dataset of 1 km terra MODIS and aqua MODIS 8-day LSTs (MOD11A2 and MYD11A2). The study time period ranged from February 2003 to January 2019. Two categorical steps were employed: (i) Assessment of LST over LULC types based on all pixels. (ii) Assessment of LST over LULC types based on meteorological station locations. Validation of LSTs was done using the air temperature of station records. Seventy station sites across Ghana were selected. The results confirm an estimation of the average [Formula: see text] consisting of [Formula: see text][Formula: see text] and [Formula: see text] RMSE, MBE, and MAE were also computed. There was a direct positive correlation between [Formula: see text] and LST over LULC types. ANOVA test method was applied to verify the accuracy of LSTs over LULC types. Croplands have the highest day LST of 39.6 °C whereas evergreen broadleaf forest has the lowest (27.0 °C). The lowest night LST was observed in savannas, grasslands, and croplands with each having 18.6 °C while the highest was water body (23.6 °C). The study revealed that an increase in elevations results in a decrease in LSTs. Also, climatic conditions and specific heat capacities of LULC types had an influence on LST estimates.

Keywords

• Land surface temperature
• air temperature
• land use land cover types
• shuttle radar topography mission
• statistical analyses
• MODIS