Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

This study presents an approach to combine uncertainties of the hydrological model outputs predicted from a number of machine learning models. The machine learning based uncertainty prediction approach is very useful for estimation of hydrological models' uncertainty in particular hydro-metrological situation in real-time application [1]. In this approach the hydrological model realizations from Monte Carlo simulations are used to build different machine learning uncertainty models to predict uncertainty (quantiles of pdf) of the a deterministic output from hydrological model . Uncertainty models are trained using antecedent precipitation and streamflows as inputs. The trained models are then employed to predict the model output uncertainty which is specific for the new input data. We used three machine learning models namely artificial neural networks, model tree, locally weighted regression to predict output uncertainties. These three models produce similar verification results, which can be improved by merging their outputs dynamically. We propose an approach to form a committee of the three models to combine their outputs. The approach is applied to estimate uncertainty of streamflows simulation from a conceptual hydrological model in the Brue catchment in UK and the Bagmati catchment in Nepal. The verification results show that merged output is better than an individual model output. [1] D. L. Shrestha, N. Kayastha, and D. P. Solomatine, and R. Price. Encapsulation of parameteric uncertainty statistics by various predictive machine learning models: MLUE method, Journal of Hydroinformatic, in press, 2013.

Distribution Of Hydrological Losses For Varying Rainfall And Antecedent Wetness Conditions

Hydrological loss is a vital component in many hydrological models, which are usedin forecasting floods and evaluating water resources for both surface and subsurface flows. Due to the complex and random nature of the rainfall runoff process, hydrological losses are not yet fully understood. Consequently, practitioners often use representative values of the losses for design applications such as rainfall-runoff modelling which has led to inaccurate quantification of water quantities in the resulting applications. The existing hydrological loss models must be revisited and modellers should be encouraged to utilise other available data sets. This study is based on three unregulated catchments situated in Mt. Lofty Ranges of South Australia (SA). The paper focuses on conceptual models for: initial loss (IL), continuing loss (CL) and proportional loss (PL) with rainfall characteristics (total rainfall (TR) and storm duration (D)), and antecedent wetness (AW) conditions. The paper introduces two methods that can be implemented to estimate IL as a function of TR, D and AW. The IL distribution patterns and parameters for the study catchments are determined using multivariate analysis and descriptive statistics. The possibility of generalising the methods and the limitations of this are also discussed. This study will yield improvements to existing loss models and will encourage practitioners to utilise multiple data sets to estimate losses, instead of using hypothetical or representative values to generalise real situations.

The Potential Of Kohonen Method For Classifying Rivers Focused On Regional Environmental Flow Determination With ELOHA

This article highlights the potential benefits that the Kohonen method has for the classification of rivers with similar characteristics by determining regional ecological flows using the ELOHA (Ecological Limits of Hydrologic Alteration) methodology. Currently, there are many methodologies for the classification of rivers, however none of them include the characteristics found in Kohonen method such as (i) providing the number of groups that actually underlie the information presented, (ii) used to make variable importance analysis, (iii) which in any case can display two-dimensional classification process, and (iv) that regardless of the parameters used in the model the clustering structure remains. In order to evaluate the potential benefits of the Kohonen method, 174 flow stations distributed along the great river basin “Magdalena-Cauca” (Colombia) were analyzed. 73 variables were obtained for the classification process in each case. Six trials were done using different combinations of variables and the results were validated against reference classification obtained by Ingfocol in 2010, whose results were also framed using ELOHA guidelines. In the process of validation it was found that two of the tested models reproduced a level higher than 80% of the reference classification with the first trial, meaning that more than 80% of the flow stations analyzed in both models formed invariant groups of streams.

RWater – A Cyber Enabled Tool For Hydrologic Modeling And Analysis

Existing distributed hydrologic models are complex and computationally demanding for using as a rapid-forecasting policy-decision tool, or even as a class-room educational tool. In addition, platform dependence, specific input/output data structures and non-dynamic data-interaction with pluggable software components inside the existing proprietary frameworks make these models restrictive only to the specialized user groups. RWater is a web-based hydrologic analysis and modeling framework that utilizes the commonly used R software within the HUBzero cyber infrastructure of Purdue University. RWater is designed as an integrated framework for distributed hydrologic simulation, along with subsequent parameter optimization and visualization schemes. RWater provides platform independent web-based interface, flexible data integration capacity, grid-based simulations, and user-extensibility. RWater uses RStudio to simulate hydrologic processes on raster based data obtained through conventional GIS pre-processing. The program integrates Shuffled Complex Evolution (SCE) algorithm for parameter optimization. Moreover, RWater enables users to produce different descriptive statistics and visualization of the outputs at different temporal resolutions. The applicability of RWater will be demonstrated by application on two watersheds in Indiana for multiple rainfall events.

Decision Support Systems And Promoting Socially Just Environmental Management

The Mauri Model DMF is unique in its approach to the management of water resources as the framework offers a transparent and inclusive approach to considering the environmental, economic, social and cultural aspects of the decisions being contemplated. The Mauri Model DMF is unique because it is capable of including multiple-worldviews and adopts mauri (intrinsic value or well-being) in the place of the more common monetised assessments of pseudo sustainability using Cost Benefit Analysis. The Mauri Model DMF uses a two stage process that first identifies participants’ worldviews and inherent bias regarding water resource management, and then facilitates transparent assessment of selected sustainability performance indicators. The assessment can then be contemplated as the separate environmental, economic, social and cultural dimensions of the decision, and collectively as an overall result; or the priorities associated with different worldviews can be applied to determine the sensitivity of the result to different cultural contexts or worldviews.

Campus Life: In Perkins Arboretum there are birds of all feathers, trees that tell a story, and even romance (especially if you're a frog)

To the uninitiated, Perkins Arboretum is the edge of campus. For those familiar with its flora and fauna, the arboretum is a wondrous classroom and refuge.

Boots on the Ground: At 40, growing Environmental Studies Program equips students to tackle real-world problems

Now 40, Colby’s Environmental Studies Program is growing exponentially, equipping students to tackle real-world problems.

Sugarcane (Saccharum X officinarum): A Reference Study for the Regulation of Genetically Modified Cultivars in Brazil

Global interest in sugarcane has increased significantly in recent years due to its economic impact on sustainable energy production. Sugarcane breeding and better agronomic practices have contributed to a huge increase in sugarcane yield in the last 30 years. Additional increases in sugarcane yield are expected to result from the use of biotechnology tools in the near future. Genetically modified (GM) sugarcane that incorporates genes to increase resistance to biotic and abiotic stresses could play a major role in achieving this goal. However, to bring GM sugarcane to the market, it is necessary to follow a regulatory process that will evaluate the environmental and health impacts of this crop. The regulatory review process is usually accomplished through a comparison of the biology and composition of the GM cultivar and a non-GM counterpart. This review intends to provide information on non-GM sugarcane biology, genetics, breeding, agronomic management, processing, products and byproducts, as well as the current technologies used to develop GM sugarcane, with the aim of assisting regulators in the decision-making process regarding the commercial release of GM sugarcane cultivars. © 2011 The Author(s).

Antler Size of Alaskan Moose Alces alces gigas: Effects of Population Density, Hunter Harvest and Use of Guides

Moose Alces alces gigas in Alaska, USA, exhibit extreme sexual dimorphism, with adult males possessing large, elaborate antlers. Antler size and conformation are influenced by age, nutrition and genetics, and these bony structures serve to establish social rank and affect mating success. Population density, combined with anthropogenic effects such as harvest, is thought to influence antler size. Antler size increased as densities of moose decreased, ostensibly a density-dependent response related to enhanced nutrition at low densities. The vegetation type where moose were harvested also affected antler size, with the largest-antlered males occupying more open habitats. Hunts with guides occurred in areas with low moose density, minimized hunter interference and increased rates of success. Such hunts harvested moose with larger antler spreads than did non-guided hunts. Knowledge and abilities allowed guides to satisfy demands of trophy hunters, who are an integral part of the Alaskan economy. Heavy harvest by humans was also associated with decreased antler size of moose, probably via a downward shift in the age structure of the population resulting in younger males with smaller antlers. Nevertheless, density-dependence was more influential than effects of harvest on age structure in determining antler size of male moose. Indeed, antlers are likely under strong sexual selection, but we demonstrate that resource availability influenced the distribution of these sexually selected characters across the landscape. We argue that understanding population density in relation to carrying capacity (K) and the age structure of males is necessary to interpret potential consequences of harvest on the genetics of moose and other large herbivores. Our results provide researchers and managers with a better understanding of variables that affect the physical condition, antler size, and perhaps the genetic composition of populations, which may be useful in managing and modeling moose populations.

Accounting for uncertainty in ecological analysis: the strengths and limitations of hierarchical statistical modeling

Analyses of ecological data should account for the uncertainty in the process(es) that generated the data. However, accounting for these uncertainties is a difficult task, since ecology is known for its complexity. Measurement and/or process errors are often the only sources of uncertainty modeled when addressing complex ecological problems, yet analyses should also account for uncertainty in sampling design, in model specification, in parameters governing the specified model, and in initial and boundary conditions. Only then can we be confident in the scientific inferences and forecasts made from an analysis. Probability and statistics provide a framework that accounts for multiple sources of uncertainty. Given the complexities of ecological studies, the hierarchical statistical model is an invaluable tool. This approach is not new in ecology, and there are many examples (both Bayesian and non-Bayesian) in the literature illustrating the benefits of this approach. In this article, we provide a baseline for concepts, notation, and methods, from which discussion on hierarchical statistical modeling in ecology can proceed. We have also planted some seeds for discussion and tried to show where the practical difficulties lie. Our thesis is that hierarchical statistical modeling is a powerful way of approaching ecological analysis in the presence of inevitable but quantifiable uncertainties, even if practical issues sometimes require pragmatic compromises.

Dogs Gone Wild: Feral Dog Damage in the United States

Feral dogs have been documented in all 50 states and estimates of damage in the U.S. from these animals amount to >$620 million annually. In Texas alone, it is estimated that over $5 million in damage to livestock annually can be attributed to feral dogs. We reviewed national statistics on feral dog damage reported to USDA, APHIS, Wildlife Services for a 10-year period from 1997 through 2006. Damage by feral dogs crossed multiple resource categories (e.g., agriculture, natural resources); some examples of damage include killing and affecting the behavior and habitat use of native wildlife; killing and maiming livestock; and their role as disease vectors to wildlife, domestic animals, and humans. We review the role of dog damage in the U.S., synthesize the amount of damage between resource categories (agriculture, human health and safety, disease, and natural resources), and report trends in dog damage during the 10-year period. Results showed an increase in dog damage across all resource categories indicating the importance of management.

Bovine Tb Eradication Project – Recognizing Hot Button Issues

The Michigan Departments of Agriculture, Community Health, and Natural Resources, US Department of Agriculture (USDA) and Michigan State University work cooperatively together as the bovine TB eradication project partners. The interagency group combines expertise in epidemiology, veterinary and human medicine, pathology, wildlife biology, animal husbandry, regulatory law and policy and risk communications. The stakeholders, those impacted by the disease, include agriculture and tourism industry representatives, “Mom-and-Pop” businesses, hunters, wildlife enthusiasts, farmers, Local Health Departments and legislators. The regulatory agencies are the above mentioned project partners, excluding MSU and USDA Wildlife Services, both of which offer services to agencies and stakeholders. Eradicating bovine TB would not be difficult if there were no social issues surrounding it. The economy, hunting traditions, animal management, tourism and human health are all impacted by regulatory response to the disease. Often the social issues play a large role in decision making, therefore it is important to understand your clientele and anticipate public reaction to policy changes and requirements.

Impacts of Plant Size, Density, Herbivory, and Desease on Native Platte Thistle (Cirsium Canescens)

Abstract. Based on prior field observations, we hypothesized that individual and interacting effects of plant size, density, insect herbivory, and especially fungal disease, influenced seedling and juvenile plant growth in native Platte thistle populations (Cirsium canescens Nutt.). We worked at Arapaho Prairie in the Nebraska Sandhills (May - August 2007), monitoring plant growth, insect damage, and fungal infection within different density thistle patches. In the main experiment, we sprayed half of test plants in different density patches with fungicide (Fungonil© Bonide, containing chlorothalonil) and half with a water control. Fungal infection rates were very low, so we found no difference in fungal attack between these treatments. However, plants that received the fungicide treatment had significantly faster growth over the season than did the control plants. At the same time, plants in the fungicide treatment had significantly reduced insect herbivory. These results strongly suggest that the fungicide had insecticidal effects and that insect herbivory significantly decreases juvenile Platte thistle growth. Further, damage by insect herbivores tended to be higher for larger plants, and herbivory was variable among different patches. However, plant density did not appear to have a large effect on the amount of insect herbivory that individual juvenile Platte thistle plants received. In the second experiment, we examined germination and survival success in relationship to seed density, and found that germination success was higher in areas of lower seed density. In the third experiment, we tested germination for filled seeds categorized primarily by color variation and size, and found no difference in germination related to either color or seed weight. We conclude that seed density, but not seed quality as estimated by color or size, affects germination success. Further, although herbivory was not significantly affected by plant density at any of the scales examined, insect herbivory significantly reduces the growth and success of juveniles of this characteristic native sand prairie plant.

Evaluating Hazelnut Cultivars for Yield, Quality and Disease Resistance

ABSTRACT This long term study focuses on testing various hazelnut cultivars for yield, nut quality and disease resistance. There are various cultivars that are being tested for these desired traits but only the Grand Traverse and Skinner will be applicable for the results of this localized study. The desired traits of commercial nut production are best matched by these two cultivars. Results from previous harvests will be used to draw trends to recommend commercially functional cultivars in Eastern Nebraska.