Predicting Phaeocystis globosa bloom in Dutch coastal waters by decision trees and nonlinear piecewise regression

Chen, Qiuwen; Mynett, Arthur E.

doi:10.1016/j.ecolmodel.2003.10.031

Cited by 35 publications

(23 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Other studies have either noted the importance of water temperature in promoting algal blooms (Watkinson et al 2005, Edwards et al 2006, Lekve et al 2006, or utilized water temperature to model algal blooms (Chen andMynett 2004, Oh et al 2007). Interestingly, however, we found no other studies to date that have focused on minimum temperatures as an important predictor of coastal algal blooms.…”

Section: Fig 2 Receiver Operating Characteristic (Roc) Curve Formentioning

confidence: 99%

Bayesian model averaging for harmful algal bloom prediction

Hamilton

McVinish

Mengersen

2009

Ecological Applications

View full text Add to dashboard Cite

Abstract. Harmful algal blooms (HABs) are a worldwide problem that have been increasing in frequency and extent over the past several decades. HABs severely damage aquatic ecosystems by destroying benthic habitat, reducing invertebrate and fish populations, and affecting larger species such as dugong that rely on seagrasses for food. Few statistical models for predicting HAB occurrences have been developed, and in common with most predictive models in ecology, those that have been developed do not fully account for uncertainties in parameters and model structure. This makes management decisions based on these predictions more risky than might be supposed. We used a probit time series model and Bayesian model averaging (BMA) to predict occurrences of blooms of Lyngbya majuscula, a toxic cyanophyte, in Deception Bay, Queensland, Australia. We found a suite of useful predictors for HAB occurrence, with temperature figuring prominently in models with the majority of posterior support, and a model consisting of the single covariate, average monthly minimum temperature, showed by far the greatest posterior support. A comparison of alternative model averaging strategies was made with one strategy using the full posterior distribution and a simpler approach that utilized the majority of the posterior distribution for predictions but with vastly fewer models. Both BMA approaches showed excellent predictive performance with little difference in their predictive capacity. Applications of BMA are still rare in ecology, particularly in management settings. This study demonstrates the power of BMA as an important management tool that is capable of high predictive performance while fully accounting for both parameter and model uncertainty.

show abstract

Section: Fig 2 Receiver Operating Characteristic (Roc) Curve Formentioning

confidence: 99%

Bayesian model averaging for harmful algal bloom prediction

Hamilton

McVinish

Mengersen

2009

Ecological Applications

View full text Add to dashboard Cite

show abstract

“…La interpretación de los resultados es muy fácil. Ha sido utilizado con éxito en temas biosanitarios, entre otros autores por Chen y Mynett (2004) 11 …”

Section: Discussionunclassified

Tuberculosis en Barcelona: modelo predictivo basado en series temporales

Muñoz¹,

Orcau²,

Caylà³

2009

Rev. Esp. Salud Publica

View full text Add to dashboard Cite

RESUMENLa tuberculosis (TB) a escala mundial sigue ocasionando cada año millones de casos nuevos y de muertes. La Organización Mundial de la Salud (OMS) ha estimado recientemente que en el 2007 se registraron 9.273.000 casos (incidencia de 139/100.000 habitantes) y 1.772.000 muertes atribuibles a esta vieja enfermedad, y en el año precedente la cifra absoluta fue ligeramente inferior (9.240.000 casos) mientras que la incidencia fue algo superior (140/100.000), por lo que se hace difí-cil precisar si hay declive o no. El objetivo de este trabajo es aplicar modelos predictivos a la TB, diferenciando entre población autóctona e inmigrante, en una ciudad en la que ha sido posible determinar el número anual de casos desde 1987.Se ha ajustado una tendencia segmentada también denominada regresión discontinua a tramos (piecewise regression o segmented regression) a las series de casos nuevos en la población autóctona e inmigrante de Barcelona. La evolución de esta enfermedad es radicalmente diferente; mientras que en la población autóctona presenta una tendencia a la baja, coincidiendo con el descenso de nuevos de casos de SIDA, en la población inmigrante, la tendencia es al alza. La estimación de nuevos casos para el años 2009 ha sido de 168 (IC 95% 109-227) y en la población inmigrante de de 227 (IC 95%, 180-275).Palabras clave: Incidencia de tuberculosis, previsión, regresión segmentada. ABSTRACT Tuberculosis in Barcelona: A Predictive model Based on Temporal SeriesTuberculosis (TB) continues to cause millions of new cases and deaths worldwide every year. The World Health Organization (WHO) has recently estimated that in 2007 9,273,000 cases (incidence 139/100,000 inhabitants) and 1,772,000 deaths were attributed to TB and, in the previous year, the absolute figure was slightly lower (9,240,000 cases) and the incidence somewhat higher (140/100,000), making it difficult to determine whether there was a reduction or not. The objective of this study was to apply fovecasting models to TB, differentiating between indigenous and immigrant subjects, in a city in which the annual number of cases has been recorded since 1987.Adjusted segmented regression (piecewise regression) was applied to the series of new cases in the indigenous and immigrant populations of Barcelona. The evolution of TB differed radically; whereas in the indigenous population there was a downward trend, coinciding with the reduction in new of cases of AIDS, in immigrants there was an upward trend. The estimated number of new cases in 2009 was 168 (95% CI 109 -227) in indigenous subjects and 227 (95% CI, 180 -275) in immigrants.

show abstract

“…telemetry (Lam, 2002)], there are only a few methods that have been employed in order to predict a HAB event. The literature found covers the following concepts:  Case-based Reasoning (CBR)  Decision trees and nonlinear piecewise regression (Chen, 2004)  Genetic programming (GP) (Muttil, 2005)  Remote sensing (RS) and marine optical sensor combinations (Sacau-Cuadrado et al, 2003;Wynne et al, 2005) Although fairly successful in their attempts to forecast HAB onset, all the aforementioned concepts were found to possess two common and substantial drawbacks: firstly, they could be too complex to be understood by anyone other than a specialist in their respective fields; and secondly, the equip-ment needed to employ them would be too costly and not very user-friendly in a universal setting. To date, nothing in the literature has been found that describes a Geographic Information System (GIS) as the basis for the generation of a HAB predictive process.…”

Section: Hab Forecast and Predictionmentioning

confidence: 99%

A Process for Harmful Algal Bloom Location Prediction Using GIS and Trend Analysis for the Terrestrial Waters of Kuwait

Shuhaibar

2008

J ENV INFORM

View full text Add to dashboard Cite

ABSTRACT. The State of Kuwait has a population of 2.5 million people, the vast majority of whom live along the southeastern stretch of its 500 km coastline. This makes the health of its coastal waters vital to the overall quality of life and economy of the country. Kuwait Bay, a shallow natural harbor situated at the eastern "mouth" of the country, lies year-round within the euphotic zone. It is exposed to many environmental stressors that originate from both point and non-point-sources. A combination of factors has led to a marked increase in harmful algal bloom (HAB) outbreaks over the past two decades. Even though they are ephemeral, these HABs have been massive and costly in environmental and economic terms. The objective of this study was to attempt to develop a process to identify potential HAB "hotspots" using GIS and trend analysis techniques, in an effort to isolate the areas that are most prone to a future HAB outbreak. ArcGIS and Crystal Ball software were used to process and analyze data provided by the Kuwait Environmental Public Authority. Data was projected over a one-year period and subsequently used to generate monthly rasters (digital images) that outlined any HAB "hotspots" present for the given month. The results showed that, although a broad snapshot of HAB-friendly conditions was produced for each month between May and October, no specificity could be applied in the context of a definite algal outbreak. Further research is needed for a more accurate process to be developed.

show abstract

Predicting Phaeocystis globosa bloom in Dutch coastal waters by decision trees and nonlinear piecewise regression

Cited by 35 publications

References 21 publications

Bayesian model averaging for harmful algal bloom prediction

Bayesian model averaging for harmful algal bloom prediction

Tuberculosis en Barcelona: modelo predictivo basado en series temporales

A Process for Harmful Algal Bloom Location Prediction Using GIS and Trend Analysis for the Terrestrial Waters of Kuwait

Contact Info

Product

Resources

About