Quantification Of Spatial Variability Research Articles

Spatial Variability Aware Deep Neural Networks (SVANN): A General Approach

Spatial variability is a prominent feature of various geographic phenomena such as climatic zones, USDA plant hardiness zones, and terrestrial habitat types (e.g., forest, grasslands, wetlands, and deserts). However, current deep learning methods follow a spatial-one-size-fits-all (OSFA) approach to train single deep neural network models that do not account for spatial variability. Quantification of spatial variability can be challenging due to the influence of many geophysical factors. In preliminary work, we proposed a spatial variability aware neural network (SVANN-I, formerly calledSVANN) approach where weights are a function of location but the neural network architecture is location independent. In this work, we explore a more flexible SVANN-E approach where neural network architecture varies across geographic locations. In addition, we provide a taxonomy of SVANN types and a physics inspired interpretation model. Experiments with aerial imagery based wetland mapping show that SVANN-I outperforms OSFA and SVANN-E performs the best of all.

Evaluation of the Use of Hyperspectral Vegetation Indices for Estimating Mangrove Leaf Area Index in Middle Andaman Island, India

ABSTRACTMangroves are highly productive ecosystems of the coastal environment in tropics and subtropics. The Leaf Area Index (LAI) is a key structural variable of the canopy, useful for estimating the primary productivity of vegetation including mangroves. We evaluated the sensitivity of narrowband spectral indices viz., Simple Ratio, Normalized Difference Vegetation Index, Soil-Adjusted Vegetation Index and Non-Linear Vegetation Index to mangrove LAI, and thereby propose the optimal bands and spectral indices for mangrove LAI estimation. The study was carried out using EO (Earth Observation)-1 Hyperion data in the Middle Andaman Island, situated in the Bay of Bengal, India. We constructed simple, normalized and non-linear band indices from all possible band pairs in the Hyperion image, and correlated them with mangrove LAI, and empirical relationships were validated through a k-fold cross-validation. The 2-D correlation plot analysis indicates that wavebands in Shortwave Infra-Red and Near Infra-Red regions, have a high potential for mangrove LAI estimation. The wavebands at 834 nm, 844 nm, 864 nm, 1034 nm, 1185 nm, 1276 nm, 1296 nm, 1629 nm and 1679 nm were the most sensitive to mangrove LAI. The non-linear vegetation indices performed better than linear indices in predicting LAI of mangrove. The identified indices are expected to improve quantification of spatial variability of mangrove LAI.

Rapid assessment of soil and contaminant variability via portable x-ray fluorescence spectroscopy: Copşa Mică, Romania

Many rural villages in Eastern Europe are severely impacted by aeolian smelter dust/deposits. Commonly, elemental studies in Romania have suffered from limited sampling numbers (n=5–10) and/or incomplete digestion offering only semi-total quantification of elements using traditional, laboratory-based techniques. These approaches are simply inadequate for evaluating potentially hazardous soils and their spatial extent, particularly at urban/rural interfaces of variable land use. Portable x-ray fluorescence (PXRF) spectroscopy can accurately quantify contamination rapidly, in-situ with a wide dynamic range and little to no sample preparation for analysis of regulated elements (e.g., As, Cd, Cu, Mn, Pb, V, Zn) and other common soil elements such as Ca, Fe, K, Rb, Sr, Ti, and Zr. A contemporary PXRF spectrometer was used to scan 61 soil samples across multiple land use types in urban/rural interfaces on-site. Each site was georeferenced with elemental data inputted into a geographic information system for high resolution kriging interpolation. These models were superimposed over modern aerial imagery to evaluate the extent of pollution for each government-regulated element with simultaneous consideration and quantification of spatial variability in naturally occurring soil elements. Pb exceeded governmental action limits across 100% of the area, while V, Mn, Cu, and Zn were exceeded in 2.2, 2.1, 39.6, and 9.8% of the area. Furthermore, many regulated elements were closely correlated to natural soil elements. In short, georeferenced PXRF data proved a powerful new tool for on-site assessment of contaminated soils; one which has rarely been utilized in Eastern Europe to date.

Tools to analyse and display variations in anatomical delineation

Variations in anatomical delineation, principally due to a combination of inter-observer contributions and image-specificity, remain one of the most significant impediments to geometrically-accurate radiotherapy. Quantification of spatial variability of the delineated contours comprising a structure can be made with a variety of metrics, and the availability of software tools to apply such metrics to data collected during inter-observer or repeat-imaging studies would allow their validation. A suite of such tools have been developed which use an Extensible Markup Language format for the exchange of delineated 3D structures with radiotherapy planning or review systems. These tools provide basic operations for manipulating and operating on individual structures and related structure sets, and for deriving statistics on spatial variations of contours that can be mapped onto the surface of a reference structure. Use of these tools on a sample dataset is demonstrated together with import and display of results in the SWAN treatment plan review system.

THE USE OF SOIL MOISTURE PROBES FOR IMPROVED UNIFORMITY AND IRRIGATION CONTROL IN GREENHOUSES

There is wide variety of soil moisture sensors available for use in greenhouse production and research applications. Such sensors can play a valuable role in improving uniformity of substrate water content in greenhouses, as well as in automating irrigation based on plant water use. Quantification of spatial variability can be used to improve the design of irrigation systems to better match plant water use. The use of soil moisture sensors for irrigation control is promising, because it can greatly reduce temporal variability in substrate water content by watering based on actual crop water use. This not only results in better temporal uniformity, but can also greatly reduce water use. Understanding the spatial distribution of substrate water content and root water uptake within a container is important in determining the optimal sensor location within a container. At the same time, the properties of different sensors need to be considered when choosing the optimal sensor for a particular application. In small containers, it may be possible to measure most of the substrate with a single sensor, while in larger containers sensors ideally would be placed in that part of the substrate where most of the water uptake occurs. INTRODUCTION High quality irrigation water is becoming increasingly scarce in many areas of the world. Population growth, increased urbanization, and drought conditions can result in a decrease in the quantity of water available for irrigation, while salt water infiltration into fresh water aquifers can affect water quality in coastal areas. Efficient irrigation practices for agricultural production, including greenhouse production, will become increasingly important. Not only is greenhouse irrigation directly affected by the quality of the available irrigation water, greenhouse irrigation practices also can affect water quality. Soilless substrates generally have a low anion exchange capacity, and greenhouse growers typically use water-soluble fertilizers. This combination can result in leaching of water and dissolved nutrients, especially nitrate and phosphate. If this leachate is not collected and recycled, runoff from greenhouses can result in significant non-point source pollution (Lea-Cox et al., 2001). Increasingly strict environmental regulations put pressure on greenhouse growers to irrigate more efficiently in order to conserve water and reduce the risk of negative environmental impacts of their facilities. To improve irrigation practices, a detailed understanding of the spatial and temporal dynamics of the water in the root zone is required. Many commercial greenhouses use timers to turn irrigation on and off. Timers do not take into account that crop water use changes on a day-to-day basis, based on environmental conditions such as solar radiation, temperature, humidity, and air circulation. For example, daily water use of vinca (Catharanthus roseus) throughout a 40day period is shown in Figure 1, along with the daily light integral (DLI, total daily a mvanier@uga.edu b sburnett@maine.edu Proc. IS on High Technology for Greenhouse Systems GreenSys2009

Spatial analysis of climate in winegrape-growing regions in Australia

Background and Aims: Temperature-based indices are commonly used to indicate long-term suitability of climate for commercially viable winegrape production of different grapevine cultivars, but their calculation has been inconsistent and often inconsiderate of within-region spatial variability. This paper (i) investigates and quantifies differences between four such indices; and (ii) quantifies the within-region spatial variability for each Australian wine region. Methods and Results: Four commonly used indices describing winegrape growing suitability were calculated for each Australian geographic indication (GI) using temperature data from 1971 to 2000. Within-region spatial variability was determined for each index using a geographic information system. The sets of indices were compared with each other using first- and second-order polynomial regression. Heat-sum temperature indices were strongly related to the simple measure of mean growing season temperature, but variation resulted in some differences between indices. Conclusion: Temperature regime differences between the same region pairs varied depending upon which index was employed. Spatial variability of the climate indices within some regions led to significant overlap with other regions; knowledge of the climate distribution provides a better understanding of the range of cultivar suitability within each region. Significance of the Study: Within-region spatial variability and the use of different indices over inconsistent time periods to describe temperature regimes have, before now, made comparisons of climates between viticulture regions difficult. Consistent calculations of indices, and quantification of spatial variability, enabled comparisons of Australian GIs to be made both within Australia and with American Viticultural Areas in the western United States.

Application of Geostatistical Tools to Quantify Spatial Variability of Selected Soil Chemical Properties from a Cultivated Tropical Peat

Abstract : Quantification of spatial variability is a vital prerequisite agriculture.This study was aimed at quantifying the spatial variability of selected chemical properties in a tropical peat cultivated with pineapple. A 1 –ha study plot was established in a commercial pineapple plantation in Simpang Rengam , Johor. Georeferenced topsoil samples ( n = 60 ) were obtained systematically from 8 x 18 m spacings in the x and y direction, respectively.These samples were tested for total C , extractable P , K , Cu , Zn and B.Soil data were first explored using univariate statistics , including normality check , non – spatial outlier detection and data transformation.This was followed by variography and kriging analyses to quantify the spatial variability of chemical properties.Results revealed a high degree of spatial variability in the majority of chemical properties , which exhibited non- normal distributions with CVs ranging from 12 to 54 %.All properties exhibited a definable spatial structure , which were described by either spherical or exponential models.Carbon , P and B showed strong spatial dependence. The majority of properties had a short effective range.Surface maps of chemical properties clearly showed spatial clustering of test values.Expecting K , all other properties showed acceptable accuracy of interpolated values. These combined data suggest the need for a site – specific approach in managing tropical peat cultivated with pineapple , particulary with regard to nutrient management.