Evaluation of Sensor and Environmental Factors Impacting the Use of Multiple Sensor Data for Time-Series Applications
AbstractMany remote sensing sensors operate in similar spatial and spectral regions, which provides an opportunity to combine the data from different sensors to increase the temporal resolution for short and long-term trend analysis. However, combining the data requires understanding the characteristics of different sensors and presents additional challenges due to their variation in operational strategies, sensor differences and environmental conditions. These differences can introduce large variability in the time-series analysis, limiting the ability to model, predict and separate real change in signal from noise. Although the research community has identified the factors that cause variations, the magnitude or the effect of these factors have not been well explored and this is due to the limitations with the real-world data, where the effects of the factors cannot be separated. Our work mitigates these shortcomings by simulating the surface, atmosphere, and sensors in a virtual environment. We modeled and characterized a deciduous forest canopy and estimated its at-sensor response for the Landsat 8 (L8) and Sentinel 2 (S2) sensors using the MODerate resolution atmospheric TRANsmission (MODTRAN) modeled atmosphere. This paper presents the methods, analysis and the sensitivity of the factors that impacts multi-sensor observations for temporal analysis. Our study finds that atmospheric compensation is necessary as the variation due to the atmosphere can introduce an uncertainty as high as 40% in the Normalized Difference Vegetation Index (NDVI) products used in change detection and time-series applications. The effect due to the differences in the Relative Spectral Response (RSR) of the two sensors, if not compensated, can introduce uncertainty as high as 20% in the NDVI products. The view angle differences between the sensors can introduce uncertainty anywhere from 9% to 40% in NDVI depending on the atmospheric compensation methods. For a difference of 5 days in acquisition, the effect of solar zenith angle can vary between 4% to 10%, depending on whether the atmospheric attenuations are compensated or not for the NDVI products. View Full-Text
Share & Cite This Article
Rengarajan, R.; Schott, J.R. Evaluation of Sensor and Environmental Factors Impacting the Use of Multiple Sensor Data for Time-Series Applications. Remote Sens. 2018, 10, 1678.
Rengarajan R, Schott JR. Evaluation of Sensor and Environmental Factors Impacting the Use of Multiple Sensor Data for Time-Series Applications. Remote Sensing. 2018; 10(11):1678.Chicago/Turabian Style
Rengarajan, Rajagopalan; Schott, John R. 2018. "Evaluation of Sensor and Environmental Factors Impacting the Use of Multiple Sensor Data for Time-Series Applications." Remote Sens. 10, no. 11: 1678.
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.