Abstract
Accurate prediction of future observations based on past data is the key to near real-time disturbance detection using satellite image time series (SITS). To overcome the limitations of existing methods, we present an attention-based long-short-term memory (LSTM) encoder-decoder model in which the historical time series of a pixel is encoded with a bidirectional LSTM encoder while the future time series is produced by another LSTM decoder. An attention mechanism is integrated into the encoder-decoder model to align the input time series with the output time series and to dynamically choose the most relevant contextual information while forecasting. Based on the proposed model, we develop a framework for near real-time disturbance detection and verify its effectiveness in the case of burned area mapping. The prediction accuracy of the proposed model is evaluated using moderate resolution imaging spectroradiometer (MODIS) time series and compared with state-of-the-art models. Experimental results show that our model achieves the best results in terms of lower prediction error and higher model fitness. We also evaluate the disturbance detection ability of the proposed framework. The proposed approach improves the detection rate of disturbances while suppressing false alarms, and increases the temporal accuracy. We suggest that the proposed methods provide new tools for enhancing current early warning systems in real time.
Highlights
T HE dramatic advances in data storage technology and high-performance computing over the past decade provide the opportunity to conduct time-series analyses with unprecedented volumes of data
Based on the proposed model, we further propose an advanced framework for near real-time disturbance detection using satellite image time series (SITS)
We investigated and compared the timeseries prediction accuracy of five models: The attention-based long-short-term memory (LSTM) encoder–decoder network, stacked LSTM network [39], artificial neural network (ANN) [38], third-order harmonic function used in change detection and classification (CCDC) [60], and season-trend model utilized in BFAST monitor [1]
Summary
T HE dramatic advances in data storage technology and high-performance computing over the past decade provide the opportunity to conduct time-series analyses with unprecedented volumes of data. Existing change detection approaches using SITS can be broadly divided into three categories, namely, window analysis methods, temporal segmentation methods, and timeseries prediction methods. From the perspective of real-time performance, time-series prediction approaches are usually superior Both window analysis methods and temporal segmentation methods need to accumulate a long series of future observations before detecting changes. An attention mechanism is integrated into the model to dynamically align historical and predicted time series In this way, the proposed model is built adaptively according to the intrinsic characteristics presented in the data, addressing the issues associated with SITS forecasting.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
