Vol 14, No 4 (2023)

Theoretical works

Multi-model ensemble successfully predicted atmospheric methane consumption in soils across the complex landscape

Glagolev M.V., Il’yasov D.V., Sabrekov A.F., Terentieva I.E., Karelin D.V.

Abstract

Methane consumption by soils is a crucial component of the CH4 and carbon cycle. It is essential to thoroughly investigate CH4 uptake by soils, particularly considering its anticipated increase by the end of the century [Zhuang et al., 2013]. Numerous mathematical models, both empirical and detailed biogeochemical [Glagolev et al., 2023], have been developed to quantify methane consumption by soils from the atmosphere. These models are instrumental in handling spatio-temporal variability and can offer reliable estimates of regional and global methane consumption by soils. Furthermore, they enhance our comprehension of the physical and biological processes that influence methanotrophy intensity. Consequently, we can forecast the response of CH4 consumption by soil to global climate shifts [Murguia-Flores et al., 2018], especially since many models consider the effects of atmospheric CH4 concentration changes on methanotrophy and ecosystem type [Zhuang et al., 2013].

In addition to the utilization of individual models, such as those cited by [Hagedorn et al., 2005; Glagolev et al., 2014; Ito et al., 2016; Silva et al., 2016], there has been extensive advancement in employing multiple models in an ensemble format. This approach aims to integrate as much a priori information as feasible [Lapko, 2002]. Throughout the 20th century, the concept of ensemble modeling evolved from merely drawing conclusions based on multiple independent experts (F. Sanders, 1963) to structured ensemble mathematical modeling [Hagedorn et al., 2005]. In this context, the term "ensemble" consistently refers to a collection containing more than one model.

Complexities in describing the physiology and biochemistry of methanotrophic bacteria in natural environments  [Bedard, Knowles, 1989; Hanson, Hanson, 1996; Belova et al., 2013; Oshkin et al., 2014] make it difficult to develop accurate biological models and determine their specific biokinetic parameters [Curry, 2007]. At the same time, broader and often empirical models, such as those by [Potter et al., 1996; Ridgwell et al., 1999; Curry, 2007; Murguia-Flores et al., 2018], demonstrate reasonable estimates of global methane consumption. Employing model ensembles could enhance accuracy, not just in global and large-scale modeling, but also at the granular level of local study sites. Nonetheless, ensemble modeling doesn't always ensure optimal outcomes, as all models within an ensemble might overlook a biological process or effect that significantly influences the dynamics of a real ecosystem [Ito et al., 2016]. For instance, no model considered anaerobic methane oxidation until this process was empirically identified [Xu et al., 2015]. Therefore, it's crucial to validate the realism of an ensemble against specific in situ data for every application. This study aimed to develop an ensemble model describing methane consumption by soils and to test its efficacy on a randomly selected study site.

In our research, we closely examined and replicated the algorithms of four soil methane consumption models: the modification by Glagolev, Filippov [2011] of Dörr et al. [1993], Curry's model [2007], the CH4 consumption block from the DLEM model [Tian et al., 2010], and the MeMo model excluding autochthonous CH4 sources [Murguia-Flores et al., 2018]. Using these, we developed an ensemble of four models. For experimental in situ data, we utilized field measurements from the Kursk region in Russia. Additionally, we introduced a method to average the ensemble model's prediction by assigning weight coefficients to each model. This approach acknowledges the idea that the total available information doubles every few years. Thus, newer models were given higher weights, while older ones received lower weights.

The model ensemble effectively predicted CH4 consumption based on in situ measurements, albeit with a notably broad confidence interval for the predictions. Notably, there was minimal variance between the standard averaging of model predictions and weighted averaging. As anticipated, individual models underperformed compared to the ensemble. We computed the Theil inconsistency coefficient for various types of means, such as quadratic mean, cubic mean, and biquadratic mean, among others [Gini, Barbensi, 1958], both for ensemble modeling results and individual models. The ensemble predictions, when averaged using diverse methods, yielded Theil inconsistency coefficients ranging from 0.156 to 0.267. The most favorable outcome (0.156) was derived from the power mean with a power index of 0.7. However, the power mean presents a challenge as its power index isn't predetermined but chosen to best fit the experimental data. A similar limitation exists for the exponential mean. While the experimental data allows for the selection of a parameter yielding a Theil coefficient of 0.157, pre-determining this optimal value (1.3) is not feasible. Regarding other estimations that don't necessitate selecting optimal parameters, it was surprising to find that one of the best results (Theil's coefficient = 0.166) came from the half-sum of extreme terms. Surprisingly, the median provided a less satisfactory result, with a Theil's coefficient of 0.222.

The merit of the ensemble approach stems from P.D. Thompson's 1977 observation, which he stated assertively: "It is an indisputable fact that two or more inaccurate, but independent predictions of the same event can be combined in such a way that their "combined" forecast, on average, will be more accurate than any of these individual forecasts" [Hagedorn et al., 2005]. Examining our ensemble of models through this lens reveals a limitation, as the condition of independence isn't fully satisfied. The models by Dörr et al. [1993], Curry [2007], and MeMo [Murguia-Flores et al., 2018] share underlying similarities and can be seen as part of a cohesive cluster. Only DLEM, crafted on entirely distinct principles, stands apart from these models. To enhance the ensemble's robustness in future iterations, the inclusion of genuinely independent models, such as a modified version of MDM [Zhuang et al., 2013] and the model by Ridgwell et al. [1999], is recommended.

The ensemble, comprising four models and implemented without specific parameter adjustments, effectively captured methane consumption across diverse sites in the Kursk region, such as fields and forests. On average, the relative simulation error for all these sites was 36%, with the experimental data displaying a variation of 26%. Notably, while the variation is modest for this dataset, methane absorption measurements generally tend to fluctuate by several tens of percent [Crill, 1991, Fig. 1; Ambus, Robertson, 2006, Fig. 3; Kleptsova et al., 2010; Glagolev et al., 2012]. Considering this broader perspective, the simulation error achieved is indeed favorable.

Upon evaluating different methods for combining individual model results within the ensemble (specifically those methods that can be applied without prior parameter adjustments based on experimental data), it was found that the most straightforward operators yielded the best outcomes. This assessment was based on Theil's inequality coefficient criterion. Both the semi-sum of extreme terms and the arithmetic mean stood out in their performance. However, a significant drawback of the constructed ensemble is the extensive confidence interval for its predictions, averaging ±78% at a 90% probability level. We hypothesize that expanding the number of independent models within the ensemble could potentially narrow this interval.

Environmental Dynamics and Global Climate Change. 2023;14(4):209-236
pages 209-236 views

Experimental works

Estimation of tree cover height on oligotrophic bog based on UAV lidar surveying

Ilyasov D.V., Kaverin A.A., Zhernov S.N., Glagolev M.V., Niyazova A.V., Kupriianova I.V., Filippov I.V., Terentieva I.E., Sabrekov A.F., Lapshina E.D.

Abstract

Выполнена верификация лидарных данных оценки высоты отдельных деревьев на участке сосново-кустарничково-сфагнового сообщества и средней высоты деревьев на 12 участках четырех типологически различных биотопах (ГМК, Рям, РмМК, Открытое болото) болотного массива «Мухрино» (средняя тайга Западной Сибири) при помощи лесотаксационных работ. Коэффициент детерминации между наземной и дистанционной оценкой высот отдельных деревьев составил 0.87, средней высоты деревьев 0.55–0.87. Представлена методика получения и обработки лидарных данных, даны методические рекомендации по увеличению точности дистанционной оценки высоты древостоя с использованием беспилотных летательных аппаратов.

Environmental Dynamics and Global Climate Change. 2023;14(4):237-248
pages 237-248 views

The influence of the hydrometeorological factors on the CO2 fluxes from the oligotrophic bog surface.

Kulik A.A., Zarov E.A.

Abstract

Global climate change is one of the most important and promising phenomena to study in actual time. One of the key causes of global climate change is increasing the greenhouse gas (GHG) concentrations in the atmosphere [IPCC, 2023]. The main greenhouse gases are methane, carbon dioxides and nitric oxide, which contribute to the greenhouse effect and global warming [Lashof, Ahuja, 1990]. Carbon dioxide (CO2) is one of the most significant and widespread gases involved in the planet's global carbon cycle [Lashof, Ahuja. 1990]. At the same time, living organisms play a key role in creation of atmosphere composition. Autotrophic organisms use a carbon dioxide to build their body structures, including complex organic compounds. During ecosystem functioning, the part of the carbon dioxide is released into the atmosphere through organism respiration, while another part is released through the decomposition of dead organic matter. Carbon dioxide may also be produced through natural and anthropogenic processes.

Peatland ecosystems play a significant role in the planet's carbon cycle, both locally and globally. Peatlands in their natural undisturbed state are a significant long-term carbon sink1. However, the process of carbon deposition is not constant in different years, peatlands may serve either as carbon sink or source2. The main factor stimulating the carbon sequestration by peatland ecosystems is climatic conditions [Harenda et al., 2018; Bond-Lamberty et al., 2018]. Peatlands are the second most significant carbon stock on Earth and the largest on land. Despite covering only 2.84% of the Earth's land surface, the amount of soil organic carbon stored in them accounts for about one-third of all soil organic carbon on Earth. Peatlands in the northern hemisphere play a particularly important role in carbon sequestration, with an estimated accumulated carbon quantity of ~473–621 Gt of carbon [Yu et al., 2010].

The largest area of peatlands in Russia is located in Western Siberia, estimated at ~42% of the total Russian area [Vomperskiy et al., 1994; Sheng et al., 2004]. The territory of Western Siberia is featured to a high share of peatlands in original undisturbed state, making them an ideal location to study the impact of global changes on peatland biogeochemical functioning worldwide.

The carbon balance of peatlands is mainly determined by two processes: photosynthesis and respiration [Harenda et al., 2018]. The main factors influencing the CO2 flux from peatlands are photosynthetically active radiation, atmospheric air temperature (Tavg), soil temperature (Tsoil), and water table level (WTL) [Miao et al., 2013; Juszczak et al., 2013; Dyukarev et al., 2019]. At the same time, the level of mutual influence and the degree of determination have not yet been fully determined.

To study the carbon balance of terrestrial ecosystems, the chamber method [Davidson et al., 2002] is widely used. The chamber method allows to estimate the CO2 flux from the surface of the ecosystem. At the same time, the use of the modern automatic system LI-COR LI-8100A (LI-COR, USA) provides high-frequency continuous data on carbon dioxide fluxes over a long period of time, which makes it possible to assess the total accumulation of carbon and significantly improve the reliability of the identified relationships with environmental factors [Zarov et al., 2022].

The purpose of this study was to assess carbon dioxide flluxes and discover the main hydrometeorological parameters that influence the flow in the hollows of the Mukhrino raised bog.

 

MATERIALS AND METHODS

The research was carried out at the «Mukhrino» field station [Dyukarev et al., 2021], located in the central part of Western Siberia, 30 km southwest of the city of Khanty-Mansiysk. The climate is featured by high repeatability of anticyclonic conditions, rapid changes in weather conditions, a humid, moderately warm summer, and a fairly harsh, snowy winter. The chamber system was installed in a homogeneous area of the peatland, dominated by Sph. balticum, C. limosa, and Scheuchzeria palustris, with the presence of E. vaginatum on the periphery. The plant composition inside the chambers was not determined, but the most homogeneous and similar areas were selected for installation (Figure 2).

Carbon dioxide flux measurements were carried out using the automated chamber method, using a portable soil respiration analysis system LI-8100A (LI-COR, USA). The flues were measured by four automated chambers installed in the raised bog area of Mukhrino (Figure 3). The first group of chambers NEE (2 LI-COR 8100-104s chambers), measured net ecosystem exchange (NEE); the second group Reco (2 LI-COR 8100-104 cameras), measured ecosystem respiration (Reco). Measurements were taken for 2 minutes every 30 minutes for all cameras. Wooden walkways were installed in the peatland area to minimize potential negative impacts on the study surface.

The fluxes were calculated using a linear model of specialized software LI-8100 File Viewer 3.0.0 (LI-COR). R programming language packages dplyr [Wickham, 2016], ggplot2 [Wickham, 2016], lubridate [Grolemund, Wickham, 2011] were used for data processing and visualization. To analyze the dynamics of NEE and Reco fluxes, the obtained values were averaged between LI-COR 8100-104s chambers (for NEE) and LI-COR 8100-104 chambers (for Reco). Gross primary production (GPP) was calculated using the equation GPP=NEE-Reco [Connolly et al., 2009]. For further analysis, measurements with a coefficient of determination (R2) of linear regression above 0.5 were selected to minimize significant noise in the data. Spearman's rank correlation method was chosen to identify dependencies of flux on hydrometeorological properties. The dependence was determined based on the data of the flux and hydrometeorological properties averaged over 30 minutes.

 

RESULTS AND DISCUSSION

The average daily variation of CO2 flows for July, September, October 2021 is shown in Fig. 5. The simultaneous use of dark and light chambers allowed to assess the flows that are released in the ecosystem as a result of the respiration of plants, animals and microorganisms (Reco), the intensity of CO2 absorption in the process photosynthesis (GPP), and net ecosystem exchange (NEE), which is the difference between the specific absorption rate (GPP) of carbon dioxide excretion (Reco). The average daily variation of Reco (Fig. 5) in July was featured by the highest values during daylight hours; the CO2 flux reaches its maximum value at 11:00 (1.44 µmol m‑2s‑1). For September and October, the daily dynamics of Reco were weakly expressed. The highest CO2 emissions were typical for evening and night time. The maximum Reco in the daily cycle was observed at 19:00 (0.47 µmol m‑2s‑1) for September, and at 00:00 (0.17 µmol m‑2s‑1) for October. The average daily cycle of GPP (Fig. 5) had a pronounced absorption maximum during daylight hours with maximum radiation, for July at 11:00 (-3.47 µmol m‑2s‑1), for September at 12:00 (-1.53 µmol m‑2s‑1), for October – at 11:00 (-0.45 µmol m‑2s‑1). The absorption of carbon dioxide from the atmosphere (GPP) had different daily durations depending on the month (Fig. 5), which is associated with a decrease in daylight hours by autumn. In July, carbon dioxide absorption was observed from 4:00 to 20:00 (16 hours), in September from 5:00 to 18:00 (13 hours), in October from 7:00 to 17:00 (10 hours). For the diurnal cycle of NEE (Fig. 5), the CO2 absorption process (GPP>Reco) predominated in the daytime, while the carbon dioxide emission process (GPPeco) dominated at night. The maximum NEE value in the daily cycle in July was estimated at 1.01 µmol m‑2s‑1 at 22:00, in September 0.49 µmol m‑2s‑1 at 20:00, in October 0.17 µmol m‑2s‑1 at 21:00. The minimum NEE value in July was -2.03 µmol m‑2s‑1 at 11:00, in September: -1.01 µmol m‑2s‑1 at 12:00, in October 0.39 µmol m‑2s‑1 at 11:00.

A total of 1711, 2625 and 1597 Reco measurements were taken in July, September and October, respectively. The highest average daily rate of ecosystem respiration Reco occurred in the third ten days of July (July 19); by the last days of October, ecosystem respiration reached its minimum in the annual course (Fig. 7). The average Reco in July was 1.05±0.25 µmol m‑2s‑1, and in October 0.13±0.01 µmol m‑2s‑1. These estimates were obtained on a sufficient array of data and therefore can be considered reliable. The peak intensity of photosynthesis was recorded on July 22, when vegetation absorbed the largest amount of CO2. After July 22, there was a gradual decline in GPP; the rate of carbon dioxide absorption in the last days of October decreased significantly, but did not drop to zero. The presence of photosynthesis in the hollow of an oligotrophic bog even in late autumn and at low air temperatures is probably due to the activity of sphagnum mosses. Net ecosystem exchange (Fig. 7) was negative every day in July, thereby the absorption of carbon dioxide from the atmosphere daily dominated its release. In September, ecosystem absorption of carbon dioxide prevailed until September 10, after which both negative and positive NEE values were observed. During this period, intense precipitation occurred, a decrease in air temperature and the amount of incoming radiation, which led to the ecosystem switching from a sink to a temporary source of CO2. In October, the number of days on which the ecosystem acted as a carbon sink decreased; on most days, carbon dioxide emissions predominated. According to average monthly values, carbon dioxide absorption prevailed in July (-0.53±0.13 µmol m‑2s‑1) and September (-0.11±0.18 µmol m‑2s‑1), in October (0.02±0.04 µmol m‑2s‑1) CO2 evolution predominated. The number of measurements according to NEE (Table 2) is greatest in September (2584) and least in July (1709).

Reco was most influenced in July (Table 3) by air and soil temperature; in September – soil temperature and marsh water level. In October, when daily temperature variability decreased, the most significant factor for Reco was PAR (-0.59). The degree of correlation of Reco with Tavg and Tsoil in July qualifies as high; these factors are directly related to Reco the higher the temperature, the greater the release of carbon dioxide into the atmosphere by the ecosystem. This is caused by an increase in the activity of microorganisms under the influence of increased temperature [Nikonova et al., 2019]. In September, the influence of Tsoil (0.81) and water level (-0.78) increased, while the influence of Tavg (0.54) decreased. The degree of correlation of these parameters with Reco in September was classified as high. It is assumed that the strong influence of water level (-0.78) on the Reco flux in September may be associated with a sharp rise in water level (Fig. 6F), which could lead to a disruption of the optimum life activity of microorganisms. Similar flow behavior was found for North American peatlands [Miao et al., 2013]. In October, the greatest influence on Reco was exerted by PAR (-0.59), the degree of correlation is weak negative; At the same time, the correlation of the indicator with PAR in July was weakly positive. The highest correlation for GPP (Table 3) was obtained with photosynthetically active radiation for all months of the study. The PAR correlation level for all months was classified as high. The inverse correlation is due to the fact that as PAR increases, CO2 absorption increases (negative GPP flux). PAR is a key factor influencing plant photosynthesis, which in turn affects their ability to assimilate CO2 and produce GPP. As PAR intensity increases, plants increase the rate of photosynthesis and absorb carbon dioxide from the atmosphere faster, which increases GPP. The greatest influence on NEE was caused by PAR (Table 3) in July, in September and October (-0.91, -0.74 and -0.71, respectively). The level of PAR correlation in July and September was high, in October it was moderate. When PAR levels increase, plants use carbon dioxide more actively to produce organic matter and increase the level of GPP in the ecosystem, which leads to an increase in NEE flux. On the other hand, when PAR levels decrease, plants become less active in photosynthesis, which leads to the prevalence of Reco and a decrease in NEE flux. Analysis of correlation coefficients calculated from data for the entire field season, the best relationship for Reco was found with soil temperature (0.88), air temperature (0.71) and water level (-0.73). PAR has the greatest influence on GPP (-0.89) and NEE (-0.73).

 

CONCLUSIONS

Automated high temporal resolution chamber measurements of carbon dioxide flux provided a data for analyzing CO2 fluxes in the peatland area. The results provided detailed information that was used to analyze the impact of environmental hydrometeorological parameters on the flux. The highest ecosystem respiration (Reco) value during a 24-hour period was recorded in July at 11:00 (1.44 µmol m‑2s‑1), in September at 19:00 (0.47 µmol m‑2s‑1), and in October at 00:00 (0.17 µmol m‑2s‑1). The maximum gross primary production (GPP) for all months occurred between 11-12 hours: in July at 11:00 (-3.47 µmol m‑2s‑1), in September at 12:00 (-1.53 µmol m‑2s‑1), and in October at 11:00 (-0.45 µmol m‑2s‑1). By autumn, the duration of GPP throughout a day decreased, as well as the amplitude of diurnal variation for all flux indicators. The highest average daily CO2 flux for all indicators was recorded in July, while the lowest was in October. In net ecosystem exchange (NEE), absorption predominated from July 14 to September 9, with days dominated by ecosystem respiration from September 10 onwards. The amplitude of the average daily flux for all indicators decreased by October.

Based on the Spearman correlation data, the highest seasonal correlation for ecosystem respiration (Reco) was with soil temperature (0.88), air temperature (0.71), and water level (-0.73). In July, the best correlation is with air temperature (0.70) and soil temperature (0.68), in September with soil temperature (0.81) and water level (-0.78), and in October with photosynthetically active radiation (PAR) (-0.59). Gross primary production (GPP) correlates best with PAR. In July, the correlation coefficient is -0.95, in September -0.86, in October -0.79, and for the entire field season -0.89. Net ecosystem exchange (NEE), similar to GPP, is most dependent on PAR. In July, the correlation coefficient is -0.91, in September -0.74, in October -0.71, and for the entire field season -0.73.

In general, the article calculates carbon dioxide fluxes from the surface of a hollow in an oligotrophic peatland. The seasonal and average daily dynamics of hydrometeorological properties are described, and their influence on CO2 flows is assessed. It is worth noting that throughout the entire growing season, the influence of external factors on fluxes decreases, reaching a minimum mutual correlation in the coldest month (October).

Environmental Dynamics and Global Climate Change. 2023;14(4):249-263
pages 249-263 views

Chronicle

VII International Field Symposium “West Siberian peatlands and the carbon cycle: past and present”

Akhmedova I.D.

Abstract

В 2024 году с 15 по 27 августа пройдет 7-й Международный полевой Симпозиум – масштабное событие для российской и мировой научной общественности, занимающейся изучением роли торфяных болот в углеродном цикле планеты, который непосредственно связан с изменением климата.

Environmental Dynamics and Global Climate Change. 2023;14(4):264-267
pages 264-267 views

This website uses cookies

You consent to our cookies if you continue to use our website.

About Cookies