Search

The recently released Landsat analysis ready data (ARD) over the United States provides the opportunity to investigate landscape dynamics using dense time series observations at 30-m resolution. However, the dataset often contains data gaps (or missing data) because of cloud contamination or data acquisition strategy, which result in different capabilities for seasonality modeling. We present a new algorithm that focuses on data gap filling using clear observations from orbit overlap regions. Multiple linear regression models were established for each pixel time series to estimate stable predictions and uncertainties. The model's training data came from stratified random samples based on the time series similarity between the pixel and data from the overlap regions. The algorithm was first evaluated using four tiles (5000 × 5000 30-m pixels for each tile) from 2018 land surface temperature data (LST) in Atlanta, Georgia. The accuracy was assessed using randomly masked clear observations with an average Root Mean Square Error (RMSE) of 3.88 and an average bias of −0.37, which were comparable to the product accuracy. We also applied the method on ARD surface reflectance bands at Fairbanks, Alaska. The accuracy assessment suggested a majority RMSE of less than 0.04 and a bias of less than 0.0023. The gap-filled time series can be of help for reliable seasonal modeling and reducing artifacts related to data availability. This approach can also be applied to other datasets, vegetation indexes, or spectral reflectance bands of other sensors. [ABSTRACT FROM AUTHOR]

Urban development and associated land cover and land use change alter the thermal, hydrological, and physical properties of the land surface. Assessments of surface urban heat island (UHI) usually focused on using remote sensing and land cover data to quantify UHI intensity and spatial distribution within a certain period. However, the mechanisms and complex interactions in landscape dynamics and land surface thermal features are still being assessed. In this study, we developed and implemented a novel approach to characterize landscape thermal conditions by focusing on UHI intensity and its spatiotemporal variation using the recently available time series of Landsat land surface temperature and land cover change products. We analyzed land surface temperature changes in urban and surrounding non-urban lands to quantify the UHI intensity and landscape thermal conditions in the Atlanta and Minneapolis metropolitan areas of the United States. Our results revealed that UHI intensities had averages of 3.4 °C and 3.3 °C in the Atlanta and Minneapolis metropolitan areas, respectively. The dominant land cover type in rural areas and urban imperviousness cover determines the UHI intensity. Increasing trends of 0.04 °C/year and 0.01 °C/year in UHI intensity between 1985 and 2018 were found in Atlanta and Minneapolis, respectively. The UHI intensity variations in 1985 and 2018 suggest that the magnitudes and temporal variations of UHI intensity averaged from all urban land cover classes are close to the UHI intensity estimated from the low intensity urban area only while the UHI intensities are more than 2 °C larger in medium to high and high intensity urban areas. The UHI intensities estimated from the maximum temperature that have statistically significant increasing trends suggest that the maximum temperature is a good element for measuring UHI effect. Urban land cover dynamics play an important role in controlling temporal variation of UHI and the UHI hotspots. Our findings support the scientific value of implementing the prototype approach as an objective framework to quantify and monitor UHI intensity at a large geographic extent. • Annual means and changes of land surface temperature in 30-m over 30 years. • Spatial distribution of urban heat island (UHI) intensity and hotspots. • Impact of land cover conditions of urban and surrounding areas on UHI intensity. • Increasing trend of UHI intensity associated with urban expansion. • Protype of UHI change analysis and monitoring in large geographic extents. [ABSTRACT FROM AUTHOR]