Mapping urban land surface temperature using remote sensing techniques and artificial neural network modelling

The objective of the present study is to monitor and predict the changes in land surface temperature (LST) in the North of Jordan during the Period 2000 to 2016. Due to political instability in the nearby countries Syria and Iraq, Jordan has witnessed increased influxes of refugees, starting from the year 2003, which has been led to the urban expansion in the area that reflected on the climatic conditions and affected the LST values. Satellite images were used for providing LST, the acquired images represented two seasons of each year, namely summer and winter. Simulation and prediction of LST values for the next 10 years were carried out using nonlinear autoregressive exogenous (NARX) artificial neural network (ANN) model. The inputs to the model consist of meteorological data collected from eight stations in the study area, population, and land use and land cover (LULC). In fact, LULC was expressed in terms of normalized difference building index (NDBI) and normalized difference vegetation index (NDVI) that were obtained from satellite images. The model showed a high correlation between these parameters and the values of simulated LST, where the correlation coefficient for the training set, validation set, testing set and for the entire data ranged from 0.91 to 0.92. Based on the predicted LST values, LST maps for the next 10 years were developed and compared with the present actual LST maps for the year 2016. The comparison has shown an average increase of 1.1 ^°C in the average LST values, which is considered a significant increase compared with previous studies.     

Mapping tropical forest structure in southeastern Madagascar using remote sensing and artificial neural networks

Tropical forest condition has important implications for biodiversity, climate change and human needs. Structural features of forests can serve as useful indicators of forest condition and have the potential to be assessed with remotely sensed imagery, which can provide quantitative information on forest ecosystems at high temporal and spatial resolutions. Herein, we investigate the utility of remote sensing for assessing, predicting and mapping two important forest structural features, stem density and basal area, in tropical, littoral forests in southeastern Madagascar. We analysed the relationships of basal area and stem density measurements to the Normalised Difference Vegetation Index (NDVI) and radiance measurements in bands 3, 4, 5 and 7 from the Landsat Enhanced Thematic Mapper Plus (ETM+). Strong relationships were identified among all of the individual bands and field based measurements of basal area (p<0.01) while there were weak and insignificant relationships among spectral response and stem density measurements. NDVI was not significantly correlated with basal area but was strongly and significantly correlated with stem density (r=−0.69, p<0.01) when using a subset of the data, which represented extreme values. We used an artificial neural network (ANN) to predict basal area from radiance values in bands 3, 4, 5 and 7 and to produce a predictive map of basal area for the entire forest landscape. The ANNs produced strong and significant relationships between predicted and actual measures of basal area using a jackknife method (r=0.79, p<0.01) and when using a larger data set (r=0.82, p<0.01). The map of predicted basal area produced by the ANN was assessed in relation to a pre-existing map of forest condition derived from a semi-quantitative field assessment. The predictive map of basal area provided finer detail on stand structural heterogeneity, captured known climatic influences on forest structure and displayed trends of basal area associated with degree of human accessibility. These findings demonstrate the utility of ANNs for integrating satellite data from the Landsat ETM+ spectral bands 3, 4, 5 and 7 with limited field survey data to assess patterns in basal area at the landscape scale.     

Regularized estimation of bathymetry and water quality using hyperspectral remote sensing

Coastal water mapping from remote-sensing hyperspectral data suffers from poor retrieval performance when the targeted parameters have little effect on subsurface reflectance, especially due to the ill-posed nature of the inversion problem. For example, depth cannot accurately be retrieved for deep water, where the bottom influence is negligible. Similarly, for very shallow water it is difficult to estimate the water quality because the subsurface reflectance is affected more by the bottom than by optically active water components.

Most methods based on radiative transfer model inversion do not consider the distribution of targeted parameters within the inversion process, thereby implicitly assuming that any parameter value in the estimation range has the same probability. In order to improve the estimation accuracy for the above limiting cases, we propose to regularize the objective functions of two estimation methods (maximum likelihood or ML, and hyperspectral optimization process exemplar, or HOPE) by introducing local prior knowledge on the parameters of interest. To do so, loss functions are introduced into ML and HOPE objective functions in order to reduce the range of parameter estimation. These loss functions can be characterized either by using prior or expert knowledge, or by inferring this knowledge from the data (thus avoiding the use of additional information).

This approach was tested both on simulated and real hyperspectral remote-sensing data. We show that the regularized objective functions are more peaked than their non-regularized counterparts when the parameter of interest has little effect on subsurface reflectance. As a result, the estimation accuracy of regularized methods is higher for these depth ranges. In particular, when evaluated on real data, these methods were able to estimate depths up to 20 m, while corresponding non-regularized methods were accurate only up to 13 m on average for the same data.

This approach thus provides a solution to deal with such difficult estimation conditions. Furthermore, because no specific framework is needed, it can be extended to any estimation method that is based on iterative optimization.     

Estimation of spatially enhanced soil moisture combining remote sensing and artificial intelligence approaches

The main objective of this study is to combine remote-sensing and artificial intelligence (AI) approaches to estimate surface soil moisture (SM) at 100 m spatial and daily temporal resolution. The two main variables used in the Triangle method, that is, land-surface temperature (LST) and vegetation cover, were downscaled and calculated at 100 m spatial resolution. LSTs were downscaled applying the Wavelet-Artificial Intelligence Fusion Approach (WAIFA) on Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat imageries. Vegetation fractions were also estimated at 100 m spatial resolution using linear spectral un-mixing and Wavelet–AI models. Vegetation indices (VIs) were replaced with the vegetation fractions obtained from sub-pixel classification in the T_s–VI triangle space. The downscaled data were then used for calculating the evaporative fraction (EF), temperature-vegetation-dryness index (TVDI), vegetation temperature condition index (VTCI), and temperature-vegetation index (TVX) at 100 m spatial resolution. Thereafter, surface SM modelling was performed using a combination of Particle Swarm Optimization with Adaptive Neuro Fuzzy Inference System (PSO-ANFIS) and Support Vector Regression (PSO-SVR) modelling approaches. Results showed that the best input data set to estimate SM includes EF, TVDI, T_s, F_vegetation, F_soil, temperature (T), precipitation at time t (P_t, P_{t – 1}, P_{t – 2}), and irrigation (I). It was also confirmed that PSO-SVR outperformed the PSO-ANFIS modelling approach and could estimate SM with a coefficient of determination (R²) of 0.93 and a root mean square error (RMSE) of 1.29 at 100 spatial resolution. Range of error was limited between ?2.64% and 2.8%. It was also shown that the method proposed by Tang et al., (2010) improved the final SM estimations.     

群智能方法在遥感信息提取中的应用分析

遥感数据作为重要的空间数据源,在众多领域发挥着不可或缺的作用。遥感信息获取技术的不断发展与遥感数据应用领域的不断扩展,促进了遥感信息提取方法的不断进步。随着人工智能算法不断被提出及成功应用,遥感信息提取领域也在逐步引入智能算法实现高效的信息提取。在对遥感信息提取方法的研究进展进行深入分析的基础上,剖析了群智能方法应用于遥感信息提取领域的潜力与优势。并应用微粒群优化方法进行遥感数据的分类,实现了微粒群优化方法应用于遥感数据分类的技术流程,取得了很好的实验结果。因此,群智能方法能够为遥感信息提取领域提供一种新的有效智能处理方法。     

Mapping multi-spectral remote sensing images using rule extraction approach

To improve the accurate rate of mapping multi-spectral remote sensing images, in this paper we construct a class of HyperRectangular Composite Neural Networks (HRCNNs), integrating the paradigms of neural networks with the rule-based approach. The supervised decision-directed learning (SDDL) algorithm is also adopted to construct a two-layer network in a sequential manner by adding hidden nodes as needed. Thus, the classification knowledge embedded in the numerical weights of trained HRCNNs can be extracted and represented in the form of If-Then rules. The rules facilitate justification on the responses to increase accuracy of the classification. A sample of remote sensing image containing forest land, river, dam area, and built-up land is used to examine the proposed approach. The accurate recognition rate reaching over 99% demonstrates that the proposed approach is capable of dealing with image mapping.     

Residential population estimation using a remote sensing derived impervious surface approach

Residential population estimation was explored based on impervious surface coverage in Marion County, Indiana, USA. The impervious surface was developed by spectral unmixing of a Landsat Enhanced Thematic Mapper (ETM+) multispectral image. The residential impervious surface was then derived by geographic information system (GIS) overlay of residential land class and impervious surface. Regression analysis was conducted to develop population density estimation models. We found that the residential impervious surface‐based approach provided the best population density estimation result, with mean and median relative errors of 38% and 23%, respectively. An overall population estimation error of ?0.97% was achieved.     

Mapping the effects of water stress on Sphagnum: Preliminary observations using airborne remote sensing

Remote sensing of near-surface hydrological conditions within northern peatlands has the potential to provide important large-scale hydrological information regarding ecological and carbon-balance processes occurring within such systems. This article details how field knowledge of the spectral properties of Sphagnum spp., airborne remote sensing data and a range of image analysis approaches, may be combined to provide a suitable proxy for near-surface wetness. Co-incident field and airborne remote sensing data were acquired in May and September 2002 over an important UK raised bog (Cors Fochno). A combination of laboratory-tested NIR and SWIR water-based and biophysical spectral reflectance indices were applied to field and airborne reflectance spectra of Sphagnum pulchrum to elucidate changes in near-surface moisture conditions. Field results showed significant correlations between water-based indices (moisture stress index (MSI) and floating water band indices (fWBI₉₈₀ and fWBI₁₂₀₀₎)) and measures of both near-surface volumetric moisture content (VMC) and water-table position. Spectral indices formulated from the NIR (fWBI₉₈₀ and fWBI₁₂₀₀) proved to be the most useful for indicating near-surface wetness across the widest range of moisture conditions because of their ability to penetrate deeper into the Sphagnum canopy. Correlations between a biophysical index based upon chlorophyll content and both hydrological measures were not significant, possibly due to relatively high levels of surface wetness at the field site in both May and September. S. pulchrum lawns were successfully located and mapped from airborne imagery using the mixed tuned match filtering (MTMF) algorithm. Importantly, MSI derived from airborne data was significantly correlated with both field moisture and the water-table position. Relationships between measures of near-surface wetness and the MSI for naturally heterogeneous canopies were, however, found to be weaker for airborne imagery than for associated field data. This is likely to be a result of the formulation of the MSI itself and the possible preferential detection of “wetter” pixels within the imagery. This effectively reduced the ability of MSI to detect subtle changes in near-surface wetness under high moisture conditions, but would not impede the use of the index under drier conditions. Results from the field data suggest that indices formulated from the NIR may be more suitable for detailed estimations of near-surface and surface wetness at the landscape-scale although reliable hyperspectral data are required to test fully the performance of such indices. The relative merits of using such an approach to determine near-surface hydrological conditions across entire peatland complexes are also discussed.     

An analytical framework for extrapolating evaporation measurements by remote sensing surface temperature

Abstract

By relating horizontal changes in evaporation to horizontal changes in surface temperature an analytical framework is formalized, by which a point measurement of evaporation can be extrapolated from a single area of uniform vegetation to a wider area of mixed vegetation. The assumptions and simplifications implicit in this technique are examined and it is shown that the relationship between the horizontal variation in evaporation and the horizontal variation in surface temperature depends strongly on the aerodynamic transfer resistance and the horizontal variation in air temperature.     

遥感边缘智能技术研究进展及挑战

随着航空航天、遥感和通信等技术的快速发展,5G等高效通信技术的革新,遥感边缘智能（edge intelligence）成为当下备受关注的研究课题。遥感边缘智能技术通过将遥感数据处理与分析技术前置实现,在近数据源的位置进行高效地遥感信息分析和决策,在卫星在轨处理解译、无人机动态实时跟踪、大规模城市环境重建和无人驾驶识别规划等应用场景中起着至关重要的作用。本文对边缘智能在遥感数据解译中的研究现状进行了归纳总结,介绍了目前遥感智能算法模型在边缘设备进行部署应用中面临的主要问题,即数据样本的限制、计算资源的限制以及灾难性遗忘问题等。针对问题具体阐述了解决思路和主要技术途径,包括小样本情况下的泛化学习方法,详细介绍了样本生成和知识复用两种解决思路;轻量化模型的设计与训练,分析了模型剪枝和量化等方法以及基于知识蒸馏的训练框架;面向多任务的持续学习方法,对比了样本数据重现和模型结构扩展两种原理。同时,还结合了典型的遥感边缘智能应用,对代表性算法的优势和不足进行了深层剖析。最后介绍了遥感边缘智能面临的挑战,以及未来技术的主要发展方向。     

Barriers to adopting satellite remote sensing for water quality management

Sustainable practices require a long-term commitment to creating solutions to environmental, social, and economic issues. The most direct way to ensure that management practices achieve sustainability is to monitor the environment. Remote sensing technology has the potential to accelerate the engagement of communities and managers in the implementation and performance of best management practices. Over the last few decades, satellite technology has allowed measurements on a global scale over long time periods, and is now proving useful in coastal waters, estuaries, lakes, and reservoirs, which are relevant to water quality managers. Comprehensive water quality climate data records have the potential to provide rapid water quality assessments, thus providing new and enhanced decision analysis methodologies and improved temporal/spatial diagnostics. To best realize the full application potential of these emerging technologies an open and effective dialogue is needed between scientists, policy makers, environmental managers, and stakeholders at the federal, state, and local levels. Results from an internal US Environmental Protection Agency qualitative survey were used to determine perceptions regarding the use of satellite remote sensing for monitoring water quality. The goal of the survey was to begin understanding why management decisions do not typically rely on satellite-derived water quality products.     

A novel model usability evaluation framework (MUsE) for explainable artificial intelligence

When it comes to complex machine learning models, commonly referred to as black boxes, understanding the underlying decision making process is crucial for domains such as healthcare and financial services, as well as when they are used in connection with safety critical systems such as autonomous vehicles. As a result, interest in explainable artificial intelligence (xAI) tools and techniques has increased in recent years. However, the user experience (UX) effectiveness of existing xAI frameworks, especially concerning algorithms that work with data as opposed to images, is still an open research question. In order to address this gap, we examine the UX effectiveness of the Local Interpretable Model-Agnostic Explanations (LIME) xAI framework, one of the most popular model agnostic frameworks found in the literature, with a specific focus on its performance in terms of making tabular models more interpretable. In particular, we apply several state of the art machine learning algorithms on a tabular dataset, and demonstrate how LIME can be used to supplement conventional performance assessment methods. Based on this experience, we evaluate the understandability of the output produced by LIME both via a usability study, involving participants who are not familiar with LIME, and its overall usability via a custom made assessment framework, called Model Usability Evaluation (MUsE), which is derived from the International Organisation for Standardisation 9241-11:2018 standard.     

Shallow water bathymetry using integrated airborne multi-spectral remote sensing

An integrated low cost airborne multi-spectral remote sensing system is described and evaluated for remote sensing for shallow water bathymetry. The system consists of: two 35mm motor driven reconnaissance cameras using colour and colour infrared film. Three optically filtered (including removable internal IR cut-off filters), electronically shuttered CCD progressive scan cameras (Sony XC-7500) integrated into an airborne direct digital recording system using a PC processor, a 32-bit RGB analogue to digital conversion card and Zip disk storage. Two CCD based imaging spectrometers providing approximately 10nm bandwidth spectral data across the CCD spectrum (400nm to 1000nm). These CCD cameras were used with a variable interference filter fixed in front of the sensor surface. This provided a 'rainbow' image of the ground varying across the image from 400nm to 700nm (visible) and 700nm to 1000nm. Field studies were undertaken to evaluate the performance of digital multi-spectral (DMV) imagery, supplementary reconnaissance photography (SRP) and VIFIS imaging spectrometry for mapping shallow water bathymetry. The results indicate good performance in shallow water and suggest that with further refinement the system could be used to give a quick comprehensive estimate of shallow water depths     

Analysis of energy fluxes and land surface parameters in a grassland ecosystem: a remote sensing perspective

A remote sensing‐based land surface characterization and flux estimation study was conducted using Landsat data from 1997 to 2003 on two grazing land experimental sites located at the Agricultural Research Services (ARS), Mandan, North Dakota. Spatially distributed surface energy fluxes [net radiation (R _n), soil heat flux (G), sensible heat (H), latent heat (LE)] and surface parameters [emissivity (ε), albedo (α), normalized difference vegetation index (NDVI) and surface temperature (T _sur)] were estimated and mapped at a pixel level from Landsat images and weather information using the Surface Energy Balance Algorithm for Land (SEBAL) procedure as a function of grazing land management: heavily grazed (HGP) and moderately grazed pastures (MGP). Energy fluxes and land surface parameters were mapped and comparisons were made between the two sites. Over the study period, H, ε and T _sur from HGP were higher by 6.7%, 18.2% and 2.9% than in MGP, respectively. The study also showed that G, LE and NDVI were higher by 1.3%, 1.6% and 7.4% for MGP than in HGP, respectively. The results of this study are beneficial in understanding the trends of land surface parameters, energy and water fluxes as a function of land management.     

Application of eigenvector analysis to remote sensing of coastal water quality

Abstract

The use of algorithms incorporating radiance information from one or a number of wavelengths is a standard technique for detecting the concentration and distribution of water quality parameters in coastal and open ocean waters. It has become clear, however, that in a turbid dynamic coastal environment there is no one algorithm applicable for all times, seasons or area because the composition of the suspended material variescontinually. Consequently site specific algorithms have been proposed. Results of an eigenvector analysis of radiance spectra and sea-truth data collected as part of airborne remote sensing campaigns in 1984 and 1985 are presented. The eigenvectors of radiance data are shown to be dependent on the type and relative concentrations of material in suspension. The technique is shown to have great potential for the identification of the composition of material in suspension without recourse to sea-truth data. This information could be used as a criterion for selection of an appropriate algorithm.     

Mapping environmental variation in lowland Amazonian rainforests using remote sensing and floristic data

This article describes a method for detailed mapping of ecological variation in a tropical rainforest based on field inventory of pteridophytes (ferns and lycophytes) and remote sensing using Landsat Enhanced Thematic Mapper Plus (ETM+) imagery. Previously known soil cation optima of the pteridophyte species were first used in calibration, i.e. to infer soil cation concentrations for sites on the basis of their pteridophyte species composition. Multiple linear regression based on spectral reflectance values in the Landsat image was then used to derive an equation that allowed the prediction of these calibrated soil values for unvisited sites in the study area. The predictive accuracy turned out to be high: the mean absolute error, as estimated by leave-one-out cross-validation, was just 7% of the total range of calibrated soil values. This method for detailed mapping of natural environmental variability in lowland tropical rainforest has applications for land-use planning, such as wildlife management, forestry, biodiversity conservation, and payments for carbon sequestration.     

Mapping natural habitats using remote sensing and sparse partial least square discriminant analysis

This work presents a novel approach for mapping the spatial distribution of natural habitats in the ‘Foothills of Larzac’ Natura 2000 listed site located in a French Mediterranean biogeographical region. Sparse partial least square discriminant analysis was used to analyse two RapidEye data sets (June 2009 and July 2010) with the purpose of choosing the most informative spectral, textural, and thematic variables that allow discrimination of habitat classes. The sparse partial least square discriminant analysis selected relevant and stable variables for the discrimination of habitat classes that could be linked to ecological or biophysical characteristics. It also gave insight into the similarities and differences between habitat classes with comparable physiognomic characteristics. The highest user accuracy was obtained for dry improved grasslands (u = 91.97%) followed by riparian ash woods (u = 88.38%). These results are very encouraging given that these two classes were identified in Annex 1 of the EC Habitats Directive as of Community interest. Due to limited data input requirements and its computational efficiency, the approach developed in this article is a good alternative to other types of variable selection approaches in a supervised classification framework and can be easily transferred to other Natura 2000 sites.     

Mapping and monitoring conifer mortality using remote sensing in the Lake Tahoe Basin

A prolonged drought in the western United States has resulted in alarming levels of mortality in conifer forests. Satellite remote sensing holds the potential for mapping and monitoring the effects of such environmental changes over large geographic areas in a timely manner. Results from the application of a forest canopy reflectance model using multitemporal Landsat TM imagery and field measurements, indicate conifer mortality can be effectively mapped and inventoried. The test area for this project is the Lake Tahoe Basin Management Unit in the Sierra Nevada of California. The Landsat TM images are from the summers of 1988 and 1991. The Li-Strahler canopy model estimates several forest stand parameters, including tree size and canopy cover for each conifer stand, from reflectance values in satellite imagery. The difference in cover estimates between the dates forms the basis for stratifying stands into mortality classes, which are used as both themes in a map and the basis of the field sampling design. Field measurements from 61 stands collected in the summer of 1992 indicate 15 % of the original timber volume in the true fir zone died between 1988 and 1992. The resulting low standard error of 11 % for this estimate indicates the utility of these mortality classes for detecting areas of high mortality. Also, the patterns in the estimated mean timber volume loss for each class follow the expected trends. The results of this project are immediately useful for fire hazard management, by providing both estimates of the degree of overall mortality and maps showing its location. They also indicate current remote sensing technology may be useful for monitoring the changes in vegetation that are expected to result from climate change.     

Mapping land use/cover distribution on a mountainous tropical island using remote sensing and GIS

The land use/cover distribution on Langkawi Island, Malaysia was mapped using remote sensing and a Geographic Information System (GIS). A Landsat Thematic Mapper (TM) satellite image taken in March 1995 was processed, geocorrected and analysed using IDRISI, raster-based GIS software. An unsupervised classification was performed based on spectral data from a composite image of the bands TM3, TM4 and TM5. Using this output, field data together with available secondary data consisting of topography, land use and soil maps were used to perform a maximum likelihood supervised classification. The overall accuracy of the output image was 90% and individual class accuracy ranged from 74% for rubber to 100% for paddy fields. The classified areas on the image were mainly confined to the mountainous and hilly regions on the island. A shaded relief map, simulating sunshine conditions, showed that the unclassified areas are located in the shadowed slopes, i.e. the slopes facing west. Consequently, the imagery was subdivided on the basis of slope aspect and a stratified classification was performed. As a result of this procedure, the overall accuracy increased to 92% and the individual class accuracy for the inland forest class increased by 9% to 90% . Using IDRISI, individual class areas as well as percentages were calculated. The kappa coefficient for the classified image was 0.90. Qualitative analysis indicates that topography is the main control on the spatial distribution of land use/cover types on the island. As Langkawi Island has been developing rapidly over the last decade, successful planning will require reliable information about land use/cover distribution and change. This study illustrates that remote sensing and GIS techniques are capable of providing such information.