The NOAA IR serves as an archival repository of NOAA-published products including scientific findings, journal articles, guidelines, recommendations, or other information authored or co-authored by NOAA or funded partners.
As a repository, the NOAA IR retains documents in their original published format to ensure public access to scientific information.
i
What influences spatial variability in restoration costs? Econometric cost models for inference and prediction in restoration planning
-
2022
-
-
Source: Biological Conservation, 274, 109710
[PDF-9.63 MB]
Details:
-
Journal Title:Biological Conservation
-
Personal Author:
-
NOAA Program & Office:
-
Description:Habitat restoration efforts are often conducted under limited budgets and information. Explicitly incorporating data on project costs can improve the effectiveness of restoration investments but gathering precise data for future projects is often infeasible at the planning stage. We demonstrate econometric and machine learning methods for improving the accessibility of cost information for restoration planning. Using geospatial matching methods, we combine 15 years of spatially explicit project records for over 1200 culvert fish passage barrier restoration worksites across the Pacific Northwest of the United States with data layers of hypothesized drivers of construction costs. We distinguish between two objectives in analyzing cost data: inference to identify cost drivers and prediction of future project costs. For inference, we multiple linear regression and find that the variables channel bankfull width and slope, road speed class, developed and agricultural land covers, and proximity to privately managed industrial land are associated with culvert restoration costs and may serve as strong proxies in a planning setting. For prediction, we use boosted regression trees to make out-of-sample projections of restoration costs for over 27,000 barrier culverts documented in state inventories. The distribution of these cost projections over space and across jurisdictions reveal higher, and more heterogeneous, culvert restoration costs in the Puget Sound region than in other nearby areas. Our results are directly applicable for resource managers making fish passage restoration decisions and serve as a template for the use of econometrics and machine learning to analyze costs recorded in restoration project databases in other contexts.
-
Source:Biological Conservation, 274, 109710
-
DOI:
-
ISSN:0006-3207
-
Format:
-
Publisher:
-
Document Type:
-
Rights Information:Accepted Manuscript
-
Compliance:Submitted
-
Main Document Checksum:
-
Download URL:
-
File Type:
