Estimating Spatial and Temporal Trends in Environmental Indices Based on Satellite Data: A Two-Step Approach

Comprehensive Insights into Spatial-Temporal Evolution Patterns, Dominant Factors of NDVI from Pixel Scale, as a Case of Shaanxi Province, China

Based on long term NDVI (1982–2015), climate, topographic factors, and land use type data information in Shaanxi Province, multiple methods (linear regression, partial and multiple correlation analysis, redundancy analysis and boosted regression trees method) were conducted to evaluate the spatial-temporal change footprints and driving mechanisms in the pixel scale. The results demonstrated that (1) the overall annual average and seasonal NDVI in this region showed a fluctuating upward trend, especially in spring. The difference between the end of season (eos) and start of season (sos) gradually increased, indicating the occurrence of temporal “greening” across most Shaanxi Province. (2) The overall spatial distribution of annual mean NDVI in Shaanxi Province was prominent in the south and low in the north, and 98.83% of the areas had a stable and increasing trend. Pixel scale analysis reflected the spatial continuity and heterogeneity of NDVI evolution. (3) Trend and breakpoint evaluation results showed that evolutionary trends were not homogeneous. There were obvious breakpoints in the latitude direction of NDVI evolution in Shaanxi Province, especially between 32–33 °N and in the north of 37 °N. (4) Compared with precipitation, the annual average temperature was significantly correlated with the vegetation indices (annual NDVI, max NDVI, time integrated NDVI) and phenology metrics (sos, eos). (5) Considering the interaction between environmental variables, the NDVI evolution was dominated by the combined influence of climate and geographic location factors in most areas.

Computational Intelligence in Remote Sensing: An Editorial

Computational intelligence is a very active and fruitful research of artificial intelligence with a broad spectrum of applications. Remote sensing data has been a salient field of application of computational intelligence algorithms, both for the exploitation of the data and for the research/development of new data analysis tools. In this editorial paper we provide the setting of the special issue "Computational Intelligence in Remote Sensing" and an overview of the published papers. The 11 accepted and published papers cover a wide spectrum of applications and computational tools that we try to summarize and put in perspective in this editorial paper.

Proximity Environmental Feature Based Tree Health Assessment Scheme Using Internet of Things and Machine Learning Algorithm

Improperly grown trees may cause huge hazards to the environment and to humans, through e.g., climate change, soil erosion, etc. A proximity environmental feature-based tree health assessment (PTA) scheme is proposed to prevent these hazards by providing guidance for early warning methods of potential poor tree health. In PTA development, tree health is defined and evaluated based on proximity environmental features (PEFs). The PEF takes into consideration the seven surrounding ambient features that strongly impact tree health. The PEFs were measured by the deployed smart sensors surrounding trees. A database composed of tree health and relative PEFs was established for further analysis. An adaptive data identifying (ADI) algorithm is applied to exclude the influence of interference factors in the database. Finally, the radial basis function (RBF) neural network (NN), a machine leaning algorithm, has been identified as the appropriate tool with which to correlate tree health and PEFs to establish the PTA algorithm. One of the salient features of PTA is that the algorithm can evaluate, and thus monitor, tree health remotely and automatically from smart sensor data by taking advantage of the well-established internet of things (IoT) network and machine learning algorithm.