scholarly journals A robust initialization method for accurate soil organic carbon simulations

2021 ◽  
Author(s):  
Eva Kanari ◽  
Lauric Cécillon ◽  
François Baudin ◽  
Hugues Clivot ◽  
Fabien Ferchaud ◽  
...  
Keyword(s):  
Machine Learning ◽  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Co2 Concentration ◽  
Monitoring Networks ◽  
Machine Learning Model ◽  
Soc Stock ◽  
Rock Eval ◽  
New Generation

Abstract. Changes in soil organic carbon (SOC) stocks are a major source of uncertainty for the evolution of atmospheric CO2 concentration during the 21st century. They are usually simulated by models dividing SOC into conceptual pools with contrasted turnover times. The lack of reliable methods to initialize these models, by correctly distributing soil carbon amongst their kinetic pools, strongly limits the accuracy of their simulations. Here, we demonstrate that PARTYsoc, a machine-learning model based on Rock-Eval® thermal analysis optimally partitions the active and stable SOC pools of AMG, a simple and well validated SOC dynamics model, accounting for effects of soil management history. Furthermore, we found that initializing the SOC pool sizes of AMG using machine-learning strongly improves its accuracy when reproducing the observed SOC dynamics in nine independent French long-term agricultural experiments. Our results indicate that multi-compartmental models of SOC dynamics combined with a robust initialization can simulate observed SOC stock changes with excellent precision. We recommend exploring their potential before a new generation of models of greater complexity becomes operational. The approach proposed here can be easily implemented on soil monitoring networks, paving the way towards precise predictions of SOC stock changes over the next decades.

scholarly journals Partitioning soil organic carbon into its centennially stable and active fractions with statistical models based on Rock-Eval® thermal analysis (PARTY<sub>SOC</sub>v2.0 and PARTY<sub>SOC</sub>v2.0<sub>EU</sub>)

2021 ◽  
Author(s):  
Lauric Cécillon ◽  
François Baudin ◽  
Claire Chenu ◽  
Bent T. Christensen ◽  
Uwe Franko ◽  
...  
Keyword(s):  
Machine Learning ◽  
Thermal Analysis ◽  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Root Mean Square Error ◽  
Mean Square ◽  
Machine Learning Model ◽  
Rock Eval ◽  
Soc Fractions

Abstract. Partitioning soil organic carbon (SOC) into two kinetically different fractions that are centennially stable or active is key information for an improved monitoring of soil health and for a more accurate modelling of the carbon cycle. However, all existing SOC fractionation methods isolate SOC fractions that are mixtures of centennially stable and active SOC. If the stable SOC fraction cannot be isolated, it has specific chemical and thermal characteristics that are quickly (ca. 1 h per sample) measureable using Rock-Eval® thermal analysis. An alternative would thus be to (1) train a machine-learning model on the Rock-Eval® thermal analysis data of soil samples from long-term experiments where the size of the centennially stable and active SOC fractions can be estimated, and (2) apply this model on the Rock-Eval® data of unknown soils, to partition SOC into its centennially stable and active fractions. Here, we significantly extend the validity range of the machine-learning model published by Cécillon et al. [Biogeosciences, 15, 2835–2849, 2018, https://doi.org/10.5194/bg-15-2835-2018], and built upon this strategy. The second version of this statistical model, which we propose to name PARTYSOC, uses six European long-term agricultural sites including a bare fallow treatment and one South American vegetation change (C4 to C3 plants) site as reference sites. The European version of the model (PARTYSOCv2.0EU) predicts the proportion of the centennially stable SOC fraction with a conservative root-mean-square error of 0.15 (relative root-mean-square error of 0.27) in a wide range of agricultural topsoils from Northwestern Europe. We plan future expansions of the PARTYSOC global model using additional reference soils developed under diverse pedoclimates and ecosystems, and we already recommend the application of PARTYSOCv2.0EU in European agricultural topsoils to provide accurate information on SOC kinetic pools partitioning that may improve the simulations of simple models of SOC dynamics.

scholarly journals Partitioning soil organic carbon into its centennially stable and active fractions with machine-learning models based on Rock-Eval® thermal analysis (PARTY<sub>SOC</sub>v2.0 and PARTY<sub>SOC</sub>v2.0<sub>EU</sub>)

2021 ◽  
Vol 14 (6) ◽  
pp. 3879-3898
Author(s):  
Lauric Cécillon ◽  
François Baudin ◽  
Claire Chenu ◽  
Bent T. Christensen ◽  
Uwe Franko ◽  
...  
Keyword(s):  
Machine Learning ◽  
Thermal Analysis ◽  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Root Mean Square Error ◽  
Mean Square ◽  
Machine Learning Model ◽  
Rock Eval ◽  
Soc Fractions

Abstract. Partitioning soil organic carbon (SOC) into two kinetically different fractions that are stable or active on a century scale is key for an improved monitoring of soil health and for more accurate models of the carbon cycle. However, all existing SOC fractionation methods isolate SOC fractions that are mixtures of centennially stable and active SOC. If the stable SOC fraction cannot be isolated, it has specific chemical and thermal characteristics that are quickly (ca. 1 h per sample) measurable using Rock-Eval® thermal analysis. An alternative would thus be to (1) train a machine-learning model on the Rock-Eval® thermal analysis data for soil samples from long-term experiments for which the size of the centennially stable and active SOC fractions can be estimated and (2) apply this model to the Rock-Eval® data for unknown soils to partition SOC into its centennially stable and active fractions. Here, we significantly extend the validity range of a previously published machine-learning model (Cécillon et al., 2018) that is built upon this strategy. The second version of this model, which we propose to name PARTYSOC, uses six European long-term agricultural sites including a bare fallow treatment and one South American vegetation change (C4 to C3 plants) site as reference sites. The European version of the model (PARTYSOCv2.0EU) predicts the proportion of the centennially stable SOC fraction with a root mean square error of 0.15 (relative root mean square error of 0.27) at six independent validation sites. More specifically, our results show that PARTYSOCv2.0EU reliably partitions SOC kinetic fractions at its northwestern European validation sites on Cambisols and Luvisols, which are the two dominant soil groups in this region. We plan future developments of the PARTYSOC global model using additional reference soils developed under diverse pedoclimates and ecosystems to further expand its domain of application while reducing its prediction error.

scholarly journals Assessing Soil Organic Carbon Stock Dynamics under Future Climate Change Scenarios in the Middle Qilian Mountains

Forests ◽  
2021 ◽  
Vol 12 (12) ◽  
pp. 1698
Author(s):  
Wei Liu ◽  
Meng Zhu ◽  
Yongge Li ◽  
Jutao Zhang ◽  
Linshan Yang ◽  
...  
Keyword(s):  
Climate Change ◽  
Machine Learning ◽  
Organic Carbon ◽  
Qilian Mountains ◽  
Future Climate ◽  
Future Climate Change ◽  
Climate Change Scenarios ◽  
Machine Learning Model ◽  
Soc Stock ◽  
The Mean

Soil organic carbon (SOC) simply cannot be managed if its amounts, changes and locations are not well known. Thus, evaluations of the spatio-temporal dynamics of SOC stock under future climate change are crucial for the adaptive management of regional carbon sequestration. Here, we evaluated the dynamics of SOC stock to a 60 cm depth in the middle Qilian Mountains (1755–5051 m a.s.l.) by combining systematic measurements from 138 sampling sites with a machine learning model. Our results reveal that the combination of systematic measurements with the machine learning model allowed spatially explicit estimates of SOC change to be made. The average SOC stock in the middle Qilian Mountains was expected to decrease under future climate change, while the size and direction of SOC stock changes seemed to be elevation-dependent. Specifically, in comparison with the 2000s, the mean annual precipitation was projected to increase by 18.37, 19.80 and 30.80 mm, and the mean annual temperature was projected to increase by 1.9, 2.4 and 2.9 °C under the Representative Concentration Pathway (RCP) 2.6 (low-emissions pathway), RCP4.5 (low-to-moderate-emissions pathway), and RCP8.5 (high-emissions pathway) scenarios by the 2050s, respectively. Accordingly, the area-weighted SOC stock and total storage for the whole study area were estimated to decrease by 0.43, 0.63 and 1.01 kg m–2 and 4.55, 6.66 and 10.62 Tg under the RCP2.6, RCP4.5 and RCP8.5 scenarios, respectively. In addition, the mid-elevation zones (3100–3900 m), especially the subalpine shrub-meadow Mollic Leptosols, were projected to experience the most intense carbon loss. However, the higher elevation zones (>3900 m), especially the alpine desert zone, were characterized by significant carbon accumulation. As for the low-elevation zones (<2900 m), SOC was projected to be less varied under future climate change scenarios. Thus, the mid-elevation zones, especially the subalpine shrub-meadows and Mollic Leptosols, should be given priority in terms of reducing CO2 emissions in the Qilian Mountains.

Space-time machine learning for modelling soil organic carbon change

2020 ◽  
Author(s):  
Gerard Heuvelink ◽  
Marcos Angelini ◽  
Laura Poggio ◽  
Zhanguo Bai ◽  
Niels Batjes ◽  
...  
Keyword(s):  
Machine Learning ◽  
Time Series ◽  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Temporal Variation ◽  
Space Time ◽  
Calibration Data ◽  
Soc Stock ◽  
Soc Stocks ◽  
Over Time

&lt;p&gt;Spatially resolved estimates of change in soil organic carbon (SOC) stocks are necessary for supporting national and international policies aimed at achieving land degradation neutrality and climate mitigation through better land management. In this work we report on the development, implementation and application of a data-driven, statistical space-time method for mapping SOC stocks, using Argentina as a pilot area. We used the Quantile Regression Forest machine-learning algorithm to predict SOC stock at 0-30 cm depth at 250 m resolution for Argentina between 1982 and 2017, on an annual basis. The model was calibrated using over 5,000 SOC stock values from the 36-year time period and 35 environmental covariates. Most covariates were static and could only explain the spatial SOC distribution. SOC change over time was modelled using time series maps of the AVHRR NDVI vegetation index. These NDVI time series maps were pre-processed using a temporal low-pass filter to allow the SOC stock for a given year to depend on the NDVI of the current as well as preceding years. Spatial patterns of SOC stock predictions were persistent over time and comparable to baseline SOC stock maps of Argentina. Predictions had modest temporal variation with an average decrease for the entire country from 2.55 kg C m&lt;sup&gt;&amp;#8209;2&lt;/sup&gt; to 2.48&amp;#160;kg&amp;#160;C&amp;#160;m&lt;sup&gt;&amp;#8209;2&lt;/sup&gt; over the 36-year period (equivalent to a decline of 210.7&amp;#160;Gg&amp;#160;C, 3.0% of the total 0&amp;#8209;30&amp;#160;cm SOC stock in Argentina). The Pampa region had a larger estimated SOC stock decrease from 4.62&amp;#160;kg&amp;#160;C&amp;#160;m&lt;sup&gt;&amp;#8209;2&lt;/sup&gt; to 4.34&amp;#160;kg&amp;#160;C&amp;#160;m&lt;sup&gt;&amp;#8209;2&lt;/sup&gt; (5.9%) during the same period. For the 2001-2015 period, predicted temporal variation was 7-fold larger than that obtained using the Tier&amp;#160;1 approach of the Intergovernmental Panel on Climate Change and the United Nations Convention to Combat Desertification. Prediction uncertainties turned out to be substantial, mainly due to the limited number and poor spatial and temporal distribution of the calibration data, and the limited explanatory power of the covariates. Cross-validation confirmed that SOC stock prediction accuracy was limited, with a Mean Error of 0.03&amp;#160;kg C m&lt;sup&gt;-2&lt;/sup&gt; and a Root Mean Squared Error of 2.04&amp;#160;kg&amp;#160;C&amp;#160;m&lt;sup&gt;-2&lt;/sup&gt;. The model explained 45% of the SOC stock variation. In spite of the large uncertainties, this work showed that machine learning methods can be used for space-time SOC mapping and may yield valuable information to land managers and policy makers, provided that SOC observation density in space and time is sufficiently large.&lt;/p&gt;

Predicting bi-decadal soil organic carbon mineralization with Rock-Eval® thermal analysis

2020 ◽  
Author(s):  
Pierre Barré ◽  
Laure Soucémarianadin ◽  
Baudin François ◽  
Chenu Claire ◽  
Bent Christensen ◽  
...  
Keyword(s):  
Thermal Analysis ◽  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Regression Models ◽  
Soil Samples ◽  
Organic C ◽  
Significant Regression ◽  
Rock Eval ◽  
Soc Mineralization

&lt;p&gt;The organic carbon reservoir of soils is a key component of climate change, calling for an accurate knowledge of the residence time of soil organic carbon (SOC). Existing proxies of the labile SOC pool such as particulate organic carbon or basal respiration tests are time consuming and unable to consistently predict SOC mineralization over years to decades. Similarly, models of SOC dynamics often yield unrealistic values of the size of SOC kinetic pools. Rock-Eval&amp;#174; 6 (RE6) thermal analysis of bulk soil samples has recently been shown to provide useful and cost-effective information regarding the long-term in-situ decomposition of SOC. The objective of this study was to design a method based on RE6 indicators to assess for a given soil, the proportion of SOC that will be mineralized in the coming 20 years.&lt;/p&gt;&lt;p&gt;To do so, we needed samples ready to be analyzed using RE6 with a known proportion of SOC mineralized in 20 years. We used archived soil samples from 4 long-term bare fallows and 8 C&lt;sub&gt;3&lt;/sub&gt;/C&lt;sub&gt;4&lt;/sub&gt; chronosequences. For each sample, the value of bi-decadal SOC mineralization was obtained from the observed SOC dynamics of its long-term bare fallow plot or the calculated C&lt;sub&gt;3&lt;/sub&gt;-derived SOC decline following the conversion to C&lt;sub&gt;4&lt;/sub&gt; plants. Those values ranged from 0.3 to 14.3 gC&amp;#183;kg&lt;sup&gt;&amp;#8722;1&lt;/sup&gt; (concentration data), representing 8.6 to 52.6% of total SOC (proportion data). All samples were analyzed using RE6 and simple linear regression models were used to predict bi-decadal SOC loss (concentration and proportion data) from 4 RE6 parameters: 1) HI (the amount of hydrogen-rich effluents formed during the pyrolysis phase of RE6; mgCH.g&lt;sup&gt;-1&lt;/sup&gt; SOC), 2) OI&lt;sub&gt;RE6&lt;/sub&gt; (the O recovered as CO and CO&lt;sub&gt;2&lt;/sub&gt; during the pyrolysis phase of RE6; mgO&lt;sub&gt;2&lt;/sub&gt;.g&lt;sup&gt;-1&lt;/sup&gt; SOC), 3) PC/SOC (the amount of organic C evolved during the pyrolysis phase of RE6; % of total SOC) and 4) T50 CO&lt;sub&gt;2&lt;/sub&gt; oxidation (the temperature at which 50% of the residual organic C was oxidized to CO&lt;sub&gt;2&lt;/sub&gt; during the RE6 oxidation phase; &amp;#176;C).&lt;/p&gt;&lt;p&gt;The RE6 HI parameter yielded the best predictions of bi-decadal SOC mineralization, for both concentration and proportion data. PC/SOC and T50 CO&lt;sub&gt;2&lt;/sub&gt; oxidation parameters also yielded significant regression models. The OI&lt;sub&gt;RE6&lt;/sub&gt; parameter was not a good predictor of bi-decadal SOC loss, with non-significant regression models. The results showed that SOC chemical composition (HI is a proxy for SOC H/C ratio), and to a lesser degree SOC thermal stability, are related to bi-decadal SOC dynamics. The RE6 thermal analysis method can therefore provide a quantitative and accurate estimate of SOC biogeochemical stability.&lt;/p&gt;

Estimation and spatio-temporal analysis of soil organic carbon stock in China using machine learning algorithms from 2000 to 2014 

2021 ◽  
Author(s):  
Yuehong Shi ◽  
Xiaolu Tang ◽  
Xinrui Luo ◽  
Zhihan Yang ◽  
Yunsen Lai ◽  
...  
Keyword(s):  
Machine Learning ◽  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Spatial Patterns ◽  
Carbon Stock ◽  
Gradient Boosting ◽  
Biogeochemical Models ◽  
Soc Stock ◽  
Soc Density ◽  
Soc Stocks

&lt;p&gt;Soil is the largest carbon pool in terrestrial ecosystems, storing up to 2 or 3 times the amount of carbon present in the atmosphere, and a small change in soil carbon stock could have profound effects on atmospheric CO&lt;sub&gt;2&lt;/sub&gt; and climate change. However, an accurate estimate of soil organic carbon (SOC) stock is still challenging. Previous studies on SOC stock prediction across China were mainly from biogeochemical models and national soil inventories, and large uncertainties still remained. In this study, we predicted SOC stock at 0 &amp;#8211; 20 cm and 0 &amp;#8211; 100 cm with 3419 and 2479field observations using artificial neural network (ANN), extreme gradient boosting (XGBoost), random forest (RF), and gradient boosting regression trees (GBRT) across China with the linkage of climate, vegetation and soil variables. Results showed that RF performed best among the four machine learning approaches with model efficiency of 0.61 for 0 &amp;#8211; 20 cm and 0.52 for 0 &amp;#8211; 100 cm. The trained RF model was used to predicted the temporal and spatial patterns of SOC stock at a spatial resolution of 1 km from 2000 to 2014 across China. Temporally, SOC stock at 0 &amp;#8211; 20 cm (p = 0.07) and 0 &amp;#8211; 100 cm (p = 0.3) did not change significantly. However, SOC density showed strong spatial patterns, the mean value of SOC density at 0-20 cm and 0-100 cm increased firstly, then decreased and then increased with the increase of latitude, and the minimum density was 39.83&amp;#176; and 41.59&amp;#176;, respectively. The total SOC stocks across China were 33.68 and 95.01 Pg C for 0 &amp;#8211; 20 cm and 0 &amp;#8211; 100 cm, respectively. The developed SOC stock could serve as an independent dataset that could be used for decision-making and help with baseline assessments for inventory and monitoring SOC stocks for global biogeochemical models in China.&lt;/p&gt;

Soil organic carbon stocks on long-term agroecosystem experiments in Canada

2010 ◽  
Vol 90 (4) ◽  
pp. 543-550 ◽  
Author(s):  
A.J. VandenBygaart ◽  
E. Bremer ◽  
B.G. McConkey ◽  
H.H. Janzen ◽  
D.A. Angers ◽  
...  
Keyword(s):  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Simulation Modelling ◽  
Western Canada ◽  
Continuous Cropping ◽  
Intergovernmental Panel ◽  
Change Factor ◽  
Soc Stock ◽  
Stock Change

Several long-term agroecosystem experiments (LTAEs) across Canada have been maintained for periods of up to a century. Much scientific knowledge of changes in soil properties through time has been learned from these few, highly productive LTAEs. We determined the effects of land management changes (LMC) on soil organic carbon (SOC) by re-sampling 27 LTAEs across Canada using identical sampling and laboratory protocols. Seven LTAEs were sampled comparing perennial to annual cropping and it was found that SOC stocks (0-30 cm) were 9.0 ± 1.5 Mg C ha-1 higher under perennial cropping after an average of 16.9 ± 2.1 yr. This yielded a SOC stock change factor of 0.6 Mg C ha-1 yr-1, comparing favourably to a modelling assessment and the Intergovernmental Panel on Climate Change (IPCC) default factor. In six LTAEs in western Canada, no-tillage increased SOC storage by 3.2 ± 1.3 Mg C ha-1 in the top 15 cm over a period of 23.3 ± 2.7 yr relative to conventional tillage, a rate of SOC storage of 0.14 Mg C ha-1 yr-1. This rate was also similar to that derived by simulation modelling and was slightly lower than the default IPCC rate for subhumid and semi-arid regions. In eastern Canada, where tillage is much deeper than western Canada, SOC storage was not significant differently between the two tillage systems. In six LTAEs in western Canada, removing fallow periods every second or third year in favour of continuous cropping increased SOC storage by 5.2 ± 1.1 Mg C ha-1 yr-1 over 21.8 ± 4.0 yr or an average SOC stock change factor of 0.23 Mg C ha-1 yr-1 to 15 cm depth. This was slightly higher than two independent meta-analyses and rates derived from simulation modelling. The results determined from a re-sampling of LTAEs across Canada provided an invaluable method of validating rates of SOC change concluded by other means.

scholarly journals Effect of Long-Term Soil Management Practices on Tree Growth, Yield and Soil Biodiversity in a High-Density Olive Agro-Ecosystem

Agronomy ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 1036
Author(s):  
Sauro Simoni ◽  
Giovanni Caruso ◽  
Nadia Vignozzi ◽  
Riccardo Gucci ◽  
Giuseppe Valboa ◽  
...  
Keyword(s):  
Organic Carbon ◽  
Soil Organic Carbon ◽  
Soil Structure ◽  
Soil Management ◽  
Carbon Pools ◽  
Grass Cover ◽  
Soil Biodiversity ◽  
Canopy Effect ◽  
Soil Organic Carbon Pools

Edaphic arthropod communities provide valuable information about the prevailing status of soil quality to improve the functionality and long-term sustainability of soil management. The study aimed at evaluating the effect of plant and grass cover on the functional biodiversity and soil characteristics in a mature olive orchard (Olea europaea L.) managed for ten years by two conservation soil managements: natural grass cover (NC) and conservation tillage (CT). The trees under CT grew and yielded more than those under NC during the period of increasing yields (years 4–7) but not when they reached full production. Soil management did not affect the tree root density. Collecting samples underneath the canopy (UC) and in the inter-row space (IR), the edaphic environment was characterized by soil structure, hydrological properties, the concentration and storage of soil organic carbon pools and the distribution of microarthropod communities. The soil organic carbon pools (total and humified) were negatively affected by minimum tillage in IR, but not UC, without a loss in fruit and oil yield. The assemblages of microarthropods benefited, firstly, from the grass cover, secondly, from the canopy effect, and thirdly, from a soil structure ensuring a high air capacity and water storage. Feeding functional groups—hemiedaphic macrosaprophages, polyphages and predators—resulted in selecting the ecotonal microenvironment between the surface and edaphic habitat.

Sign in / Sign up

Export Citation Format

Share Document