Batch-sequential design and heteroskedastic surrogate modeling for delta smelt conservation

Zhang, Boya; Gramacy, Robert B.; Johnson, Leah R.; Rose, Kenneth A.; Smith, Eric

doi:10.1214/21-aoas1521

Cited by 3 publications

(3 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The realized growth rate sensitivity analysis indicated that spring recruitment and summer survival, and the underlying covariates controlling their deterministic values, were the most influential factors affecting Delta Smelt population growth rates. The importance of summertime processes, particularly Delta Smelt body condition as affected by food, agrees with a model sensitivity analysis by Zhang et al (2022) of an individual based life cycle model developed by Rose et al (2013). Similarly, greater changes to the growth rate due to changes in summer outflows in comparison to changes in winter south Delta OMR flows were found by Smith et al (2021).…”

Section: Discussionsupporting

confidence: 85%

“…If so, increasing outflow in June might increase the frequency of years that the Delta Smelt population grows. However, a recent modeling experiment coupling Delta Smelt bioenergetics with other habitat suitability metrics suggests that foraging conditions deteriorate as the summer progresses and that higher outflows help better align the Delta Smelt's spatial distribution with the best remaining foraging habitat conditions (Smith & Nobriga, 2023; see also Kimmerer & Rose, 2018; Zhang et al, 2022). This implies that outflow after June may be the mechanistic driver of Delta Smelt survival, or that extended periods of very elevated outflow in the summer are necessary for robust population growth given the very altered ecological context in which Delta Smelt now live.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Identifying minimum freshwater habitat conditions for an endangered fish using life cycle analysis

Polansky,

Mitchell,

Nobriga

2024

Conservat Sci and Prac

View full text Add to dashboard Cite

Identifying the most important factors affecting population growth in animal life cycles is an important tool of species conservation. Delta Smelt (Hypomesus transpacificus), an annual fish endemic to the San Francisco Estuary in California (USA), has been provided legal protection since 1993 but 30 years later exists in a conservation‐reliant state on the brink of extinction. Despite considerable controversies about what factors are most responsible for the species' decline, no population growth rate sensitivity comparisons between the most important factors regulating growth have been done. Nor has anyone attempted to quantitatively identify habitat conditions needed to support positive population growth. We developed a set of stage‐structured population models to link habitat indices regulating recruitment of new generations of fish as they metamorphosed into juveniles and the subsequent survival of those fish over several seasons until they reached adulthood. These models are used to quantify drivers of growth rate variation over 30 years. Several complimentary sensitivity analyses indicated freshwater outflow to the estuary during summer had the largest potential to change population growth. Multiple habitat metrics (e.g., food availability, temperature) influencing recruitment and life stage specific survival rates across different seasons interacted in nonlinear ways to determine habitat conditions and water management targets associated with positive population growth. We discuss the implications for freshwater management, Delta Smelt conservation, and the challenges climate change poses for co‐implementation of these two societal priorities.

show abstract

Section: Discussionsupporting

confidence: 85%

Section: Discussionmentioning

confidence: 99%

Identifying minimum freshwater habitat conditions for an endangered fish using life cycle analysis

Polansky,

Mitchell,

Nobriga

2024

Conservat Sci and Prac

View full text Add to dashboard Cite

show abstract

“…We have found that it is beneficial to generate a representative downsample of the original data using methods such as Mak and Joseph [22]; Vakayil and Joseph [23] when generating the KDE weight function. Finally, we note that the approach above can be further extended to a non‐uniform analog of the batch sequential framework as described in Zhang et al [24].…”

Section: Non‐uniform Active Learning For Gaussian Processesmentioning

confidence: 99%

Non‐uniform active learning for Gaussian process models with applications to trajectory informed aerodynamic databases

Quinlan,

Movva,

Perfect

2024

Statistical Analysis

View full text Add to dashboard Cite

The ability to non‐uniformly weight the input space is desirable for many applications, and has been explored for space‐filling approaches. Increased interests in linking models, such as in a digital twinning framework, increases the need for sampling emulators where they are most likely to be evaluated. In particular, we apply non‐uniform sampling methods for the construction of aerodynamic databases. This paper combines non‐uniform weighting with active learning for Gaussian Processes (GPs) to develop a closed‐form solution to a non‐uniform active learning criterion. We accomplish this by utilizing a kernel density estimator as the weight function. We demonstrate the need and efficacy of this approach with an atmospheric entry example that accounts for both model uncertainty as well as the practical state space of the vehicle, as determined by forward modeling within the active learning loop.

show abstract

A Variance-Based Sensitivity Analysis Approach for Identifying Interactive Exposures

Lu,

Zhang,

Birukov

et al. 2024

Stat Biosci

View full text Add to dashboard Cite

Batch-sequential design and heteroskedastic surrogate modeling for delta smelt conservation

Cited by 3 publications

References 51 publications

Identifying minimum freshwater habitat conditions for an endangered fish using life cycle analysis

Identifying minimum freshwater habitat conditions for an endangered fish using life cycle analysis

Non‐uniform active learning for Gaussian process models with applications to trajectory informed aerodynamic databases

A Variance-Based Sensitivity Analysis Approach for Identifying Interactive Exposures

Contact Info

Product

Resources

About