4.7 Article

Machine learning models inaccurately predict current and future high-latitude C balances

相关参考文献

注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article Environmental Sciences

Rapidly changing high-latitude seasonality: implications for the 21st century carbon cycle in Alaska

Ian A. Shirley et al.

Summary: This study examines how climate warming will shift carbon cycle seasonality in Alaska throughout the 21st century, using a well-tested ecosystem model. The results show that spring net carbon uptake will overtake summer net carbon uptake by year 2100, driven by a relaxation of temperature limitation on plant productivity. Additionally, warmer soil temperatures and increased carbon inputs from plants lead to larger carbon losses in fall and winter compared to summer net uptake. However, the large increases in spring net carbon uptake suggest that the high-latitude atmospheric carbon sink will sustain throughout this century.

ENVIRONMENTAL RESEARCH LETTERS (2022)

Article Multidisciplinary Sciences

Inferring causal relations from observational long-term carbon and water fluxes records

Emiliano Diaz et al.

Summary: This paper introduces a methodology based on causal discovery methods to infer spatial patterns of causal relations between key variables of carbon and water cycles. The proposed methodology addresses the issues of noise levels and variable coupling in convergent cross-mapping (CCM) and combines temporal bootstrapping decision scores with information-geometric causal inference (IGCI) to derive robust and stringent cross-map skill scores.

SCIENTIFIC REPORTS (2022)

Article Ecology

Modeling terrestrial net ecosystem exchange using machine learning techniques based on flux tower measurements

Hassan Abbasian et al.

Summary: Identifying the complex relationships between Net Ecosystem Exchange (NEE) of CO2 and land-atmosphere interactions is crucial for understanding the dynamic of carbon fluxes. This study used various machine learning methods to predict daily NEE magnitude and evaluated the performance of the models using statistical indices. The results showed variations in model performance across different plant functional types and phenological periods.

ECOLOGICAL MODELLING (2022)

Article Environmental Sciences

Upscaling Net Ecosystem Exchange Over Heterogeneous Landscapes With Machine Learning

O. Reitz et al.

Summary: This study explores different feature selection methods and CO2 flux datasets to predict daily CO2 fluxes using a Random Forest model. While backward feature elimination was effective in removing irrelevant variables, high-quality-low-quantity flux datasets were important for improving predictions. Spatial prediction was more challenging than temporal prediction, with the model underestimating the variance of CO2 fluxes. Despite difficulties, the Random Forest model showed promise for predicting CO2 fluxes at a high spatial resolution.

JOURNAL OF GEOPHYSICAL RESEARCH-BIOGEOSCIENCES (2021)

Article Environmental Sciences

Topographical Controls on Hillslope-Scale Hydrology Drive Shrub Distributions on the Seward Peninsula, Alaska

Zelalem A. Mekonnen et al.

Summary: Observations suggest that shrubs are expanding in the Arctic tundra due to climate warming, with topography playing a key role in determining shrub expansion by influencing water and nutrient dynamics. The study highlights the importance of considering hillslope-scale coupled surface and subsurface hydrology in land models to accurately predict current plant distributions and future trajectories in Arctic ecosystems.

JOURNAL OF GEOPHYSICAL RESEARCH-BIOGEOSCIENCES (2021)

Article Biodiversity Conservation

Greening of the earth does not compensate for rising soil heterotrophic respiration under climate change

Dilip G. T. Naidu et al.

Summary: The study evaluates the impact of temperature and precipitation changes on NPP and Rh using global time-series datasets, finding that changes in C-influx may not compensate for C-efflux in wetter and warmer conditions. The vulnerability of soil-C is primarily determined by precipitation, suggesting that soil-C is vulnerable to climate change in a warmer world. There remains uncertainty in the impact of climate change on soil-C stewardship.

GLOBAL CHANGE BIOLOGY (2021)

Article Multidisciplinary Sciences

Spatial heterogeneity and environmental predictors of permafrost region soil organic carbon stocks

Umakant Mishra et al.

Summary: Large stocks of organic carbon have accumulated in the permafrost region of the Northern Hemisphere, with estimated amounts in the top 3 meters of soil. Analysis of over 2700 soil profiles revealed that soil wetness index and elevation are key topographic controllers, while surface air temperature and precipitation play significant roles in determining SOC stocks. The study's high-resolution geospatial assessment provides important insights for modeling the response of permafrost-affected soils to climate change.

SCIENCE ADVANCES (2021)

Review Environmental Sciences

Arctic tundra shrubification: a review of mechanisms and impacts on ecosystem carbon balance

Zelalem A. Mekonnen et al.

Summary: Shrub expansion in the Arctic tundra is a significant response to rapid climate warming, impacting ecosystem carbon balances through various factors such as climate change, accelerated nutrient cycling, and changing disturbance regimes. The increase in tall deciduous shrubs can alter carbon dynamics by enhancing carbon uptake, affecting ecosystem respiration, and influencing feedback mechanisms that impact snowpack, permafrost, energy balance, and litter inputs. Land models need to incorporate plant traits to accurately predict tundra vegetation dynamics and carbon balances over decadal to centennial scales.

ENVIRONMENTAL RESEARCH LETTERS (2021)

Article Biodiversity Conservation

Statistical upscaling of ecosystem CO2 fluxes across the terrestrial tundra and boreal domain: Regional patterns and uncertainties

Anna-Maria Virkkala et al.

Summary: The study analyzed the spatial patterns and drivers of CO2 fluxes in high-latitude regions using statistical models, showing that machine learning and ensemble predictions outperformed traditional regression methods. CO2 sink strength was found to be larger in boreal biome compared to tundra, with the variability reflecting heterogeneity in soil organic carbon stocks, climate, and vegetation productivity. Despite uncertainty, the high-latitude region was estimated to be an annual CO2 sink during 1990-2015 based on ensemble predictions of NEE.

GLOBAL CHANGE BIOLOGY (2021)

Article Environmental Sciences

Non-growing season plant nutrient uptake controls Arctic tundra vegetation composition under future climate

William J. Riley et al.

Summary: Plant growth and distribution in high-latitude tundra ecosystems are limited by nutrient availability, and the acquisition of nutrients by plants during the non-growing season significantly affects vegetation dynamics. Utilizing a mechanistic model, this study demonstrates that nutrient uptake during the non-growing season enhances shrub growth and expansion rates.

ENVIRONMENTAL RESEARCH LETTERS (2021)

Article Computer Science, Artificial Intelligence

Shortcut learning in deep neural networks

Robert Geirhos et al.

NATURE MACHINE INTELLIGENCE (2020)

Article Environmental Sciences

Evaluating temporal controls on greenhouse gas (GHG) fluxes in an Arctic tundra environment: An entropy-based approach

Bhavna Arora et al.

SCIENCE OF THE TOTAL ENVIRONMENT (2019)

Article Multidisciplinary Sciences

The global tree restoration potential

Jean-Francois Bastin et al.

SCIENCE (2019)

Article Environmental Sciences

Large loss of CO2 in winter observed across the northern permafrost region

Susan M. Natali et al.

NATURE CLIMATE CHANGE (2019)

Review Geosciences, Multidisciplinary

Monthly gridded data product of northern wetland methane emissions based on upscaling eddy covariance observations

Olli Peltola et al.

EARTH SYSTEM SCIENCE DATA (2019)

Article Environmental Sciences

Machine learning methods for crop yield prediction and climate change impact assessment in agriculture

Andrew Crane-Droesch

ENVIRONMENTAL RESEARCH LETTERS (2018)

Article Multidisciplinary Sciences

Detecting the Causal Effect of Soil Moisture on Precipitation Using Convergent Cross Mapping

Yunqian Wang et al.

SCIENTIFIC REPORTS (2018)

Article Multidisciplinary Sciences

Carbon dioxide sources from Alaska driven by increasing early winter respiration from Arctic tundra

Roisin Commane et al.

PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA (2017)

Article Environmental Sciences

Lightning as a major driver of recent large fire years in North American boreal forests

Sander Veraverbeke et al.

NATURE CLIMATE CHANGE (2017)

Article Plant Sciences

Ground ice melt in the high Arctic leads to greater ecological heterogeneity

Michael S. Becker et al.

JOURNAL OF ECOLOGY (2016)

Article Geography, Physical

Scaling-up permafrost thermal measurements in western Alaska using an ecotype approach

William L. Cable et al.

CRYOSPHERE (2016)

Article Environmental Sciences

Causal feedbacks in climate change

Egbert H. van Nes et al.

NATURE CLIMATE CHANGE (2015)

Article Ecology

Representativeness-based sampling network design for the State of Alaska

Forrest M. Hoffman et al.

LANDSCAPE ECOLOGY (2013)

Article Environmental Sciences

Shifts in Arctic vegetation and associated feedbacks under climate change

Richard G. Pearson et al.

NATURE CLIMATE CHANGE (2013)

Article Biodiversity Conservation

A frozen feast: thawing permafrost increases plant-available nitrogen in subarctic peatlands

Frida Keuper et al.

GLOBAL CHANGE BIOLOGY (2012)

Article Multidisciplinary Sciences

Detecting Causality in Complex Ecosystems

George Sugihara et al.

SCIENCE (2012)

Article Geography, Physical

The emergence of surface-based Arctic amplification

M. C. Serreze et al.

CRYOSPHERE (2009)

Article Ecology

A working guide to boosted regression trees

J. Elith et al.

JOURNAL OF ANIMAL ECOLOGY (2008)

Article Computer Science, Information Systems

Function matching for Soviet-era table-based modified polyconic projections

IO Bildirici et al.

INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE (2006)

Article Computer Science, Information Systems

Efficient regionalization techniques for socio-economic geographical units using minimum spanning trees

R. M. Assuncao et al.

INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE (2006)

Article Biodiversity Conservation

The evidence for shrub expansion in Northern Alaska and the Pan-Arctic

K Tape et al.

GLOBAL CHANGE BIOLOGY (2006)