Estimating uncertainty in density surface models.

Area of Science:

• Ecology
• Marine Biology
• Statistical Modeling

Background:

• Species distribution models (SDMs) are vital for ecological management and conservation.
• Quantifying uncertainty in SDM predictions is essential but often challenging.
• Existing guidance on managing and combining uncertainty sources in SDMs is limited.

Purpose of the Study:

• To identify and characterize sources of uncertainty in density surface models (DSMs).
• To propose a flexible and modular approach for encapsulating these uncertainties.
• To demonstrate the application of this approach using cetacean density modeling.

Main Methods:

• Focus on density surface models, a spatial modeling technique for distance sampling data.
• Development of an extensible, modular, hybrid analytical-simulation framework.
• Application to fin whale (Balaenoptera physalus) density estimation in the California Current Ecosystem.

Main Results:

• Detailed identification of where uncertainty can arise within the multi-stage DSM process.
• Demonstration of a novel hybrid approach to systematically capture and combine uncertainty.
• Empirical analysis of fin whale distribution and associated prediction uncertainty.

Conclusions:

• The proposed hybrid approach offers a robust method for uncertainty quantification in DSMs.
• This framework enhances the reliability of SDM predictions for marine mammal management.
• Improved understanding of uncertainty aids conservation efforts for species like fin whales.