Misclassification error in satellite imagery data: Implications for empirical land-use models

Sandler, Austin; Rashford, Benjamin S.

doi:10.1016/j.landusepol.2018.04.008

Cited by 7 publications

(5 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Second, we assumed the land-cover map provided an accurate assessment of the proportion of spring-seeded crops and grassland in the landscapes. Errors with remote sensing can induce biased estimates, such that percentage of crops in the landscape is underestimated and grassland is overestimated in the PPR 56 . If present, this bias would make pintail persistence even less likely than in our current predictions.…”

Section: Discussionmentioning

confidence: 99%

Linking landscape-scale conservation to regional and continental outcomes for a migratory species

Mattsson

Devries

Dubovsky

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Land-use intensification on arable land is expanding and posing a threat to biodiversity and ecosystem services worldwide. We develop methods to link funding for avian breeding habitat conservation and management at landscape scales to equilibrium abundance of a migratory species at the continental scale. We apply this novel approach to a harvested bird valued by birders and hunters in north America, the northern pintail duck (Anas acuta), a species well below its population goal. Based on empirical observations from 2007-2016, habitat conservation investments for waterfowl cost $313 M and affected <2% of the pintail's primary breeding area in the Prairie Pothole Region of Canada. Realistic scenarios for harvest and habitat conservation costing an estimated $588 M (2016 USD) led to predicted pintail population sizes <3 M when assuming average parameter values. Accounting for parameter uncertainty, converting 70-100% of these croplands to idle grassland (cost: $35.7B-50B) is required to achieve the continental population goal of 4 M individuals under the current harvest policy. Using our work as a starting point, we propose continued development of modeling approaches that link conservation funding, habitat delivery, and population response to better integrate conservation efforts and harvest management of economically important migratory species. The expansion of intensive agriculture threatens the diversity of species 1-4 and ecosystem services 5 associated with arable landscapes. Growing demands for food, feed, fuel, and fiber have precipitated such land-use intensification in these and other landscapes worldwide 6,7. Substantial evidence shows that people benefit from high levels of biodiversity through many associated ecosystem services including food security, water quality, and recreation 8,9. Birds are valued by the general public and recreationists, including birders and hunters in the case of game species 10-12. Some grassland and farmland birds have undergone especially steep declines tied to intensifying agricultural practices 13. Many of these populations comprise individuals undertaking annual long-distance movements between breeding and wintering grounds often spanning multiple countries 14,15. International funds for conserving critical habitats at the landscape scale in North America are essential for maintaining these populations 16,17 and these resources vary over time 18. This conservation challenge begs the question of how shifts in funding and resultant changes in landscape composition might affect migratory birds. Important advances have been made in modeling impacts of habitat changes on migratory birds at regional and continental scales. Existing approaches have important limitations, however. Some lack landscape-level scenarios for habitat conservation 19-21 , while others focus on hypothetical species 22. Case studies most commonly lack predicted changes in species-specific abundance, predicting instead shifts in community-level metrics 23,24 or indices of habitat suitability 25-27. Spec...

show abstract

Section: Discussionmentioning

confidence: 99%

Linking landscape-scale conservation to regional and continental outcomes for a migratory species

Mattsson

Devries

Dubovsky

et al. 2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Therefore, it is strictly suggested to dispense the correct classified map to the algorithm. Moreover, many efforts have been made by the researcher to resolve the misclassified (it means when a class category is assigned to a different category instead of the actual one) issue in PCC with the help of Bayesian soft fusion 31 , 64 . Here ANN and PCC named ANPC have been integrated to detect the multitemporal changes over agricultural land.…”

Section: Methodsmentioning

confidence: 99%

“…Moreover, many efforts have been made by the researcher to resolve the misclassified (it means when a class category is assigned to a different category instead of the actual one) issue in PCC with the help of Bayesian soft fusion. 31,64 Here ANN and PCC named ANPC have been integrated to detect the multitemporal changes over agricultural land. For cross-referencing, a comparative analysis is performed with respect to well-known PCC-based change detection methods, i.e., RFPC and SVMPC.…”

Section: Change Detectionmentioning

confidence: 99%

Qualitative and quantitative analysis of artificial neural network-based post-classification comparison to detect the earth surface variations using hyperspectral and multispectral datasets

Dahiya

Gupta

Singh

2023

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

Remote sensing is an effective way to analyze land surface changes on regular basis globally. In the previous literature, numerous change detection models were developed to detect the multitemporal or seasonal variations using different optical datasets, such as multispectral and hyperspectral. But there are many challenges involved in the processing of numerous bands, especially in the case of hyperspectral datasets, such as computational constraints and radiometric/atmospheric distortion. A simple framework-based artificial neural network (ANN) and postclassification comparison (PCC), named ANN-based PCC (ANPC), has been proposed to detect the multitemporal changes over agricultural land using hyperspectral, i.e., Earth Observation (EO-1) Hyperion and multispectral, i.e., Landsat-8. For cross-referencing, a comparative analysis is performed with respect to well-known PCC-based change detection methods, i.e., random forest-based post-comparison (RFPC) and support vector machine-based post-comparison (SVMPC). Experimental outcomes confirmed the effectiveness of ANPC (with an accuracy of more than 90%) in the extraction of multitemporal changes as compared to RFPC and SVMPC (with an accuracy of <90%). Our study enhances the utilization of the hyperspectral dataset (due to narrow spectral bands) in the extraction of critical information about the earth's surface parameters.

show abstract

“…Given the model specification complexity we did not control for potential misclassification error in the outcome variable, which may cause attenuated coefficient estimates [93,94]. Exploratory analysis suggests the accuracy of the observed wasted children is as low as 37% (Nigeria) and 21% (Kenya).…”

Section: Limitationsmentioning

confidence: 99%

The Socio-Environmental Determinants of Childhood Malnutrition: A Spatial and Hierarchical Analysis

Sandler,

Sun

2024

Nutrients

View full text Add to dashboard Cite

Despite a remarkable reduction in global poverty and famines, substantial childhood malnutrition continues to persist. In 2017, over 50 million and 150 million young children suffered from acute malnutrition (wasting) and chronic malnutrition (stunting), respectively. Yet, the measurable impact of determinants is obscure. We evaluate proposed socio-environmental related determinants of stunting and wasting across Kenya and Nigeria and quantify their effectiveness. We combine health and demographic data from Kenya and Nigeria Demographic Health Surveys (2003, 2008–2009, 2013, 2014) with spatially explicit precipitation, temperature, and vegetation data. Geospatial and disaggregated data help to understand better who is at risk and where to target mitigation efforts. We evaluate the responsiveness of malnutrition indicators using a four-level random intercept hierarchical generalized logit model. We find that spatial and hierarchical relationships explain 28% to 36% of malnutrition outcome variation. Temporal variation in precipitation, temperature, and vegetation corresponds with more than a 50% change in malnutrition rates. Wasting is most impacted by mother’s education, family wealth, clinical delivery, and vaccinations. Stunting is most impacted by family wealth, mother’s education, clinical delivery, vaccinations, and children asymptomatic of fever, cough, or diarrhea. Remotely monitored climatic variables are powerful determinants, however, their effects are inconsistent across different indicators and locations.

show abstract

Misclassification error in satellite imagery data: Implications for empirical land-use models

Cited by 7 publications

References 78 publications

Linking landscape-scale conservation to regional and continental outcomes for a migratory species

Linking landscape-scale conservation to regional and continental outcomes for a migratory species

Qualitative and quantitative analysis of artificial neural network-based post-classification comparison to detect the earth surface variations using hyperspectral and multispectral datasets

The Socio-Environmental Determinants of Childhood Malnutrition: A Spatial and Hierarchical Analysis

Contact Info

Product

Resources

About