Empirical Regression Models Research Articles

A new method for estimating carbon dioxide emissions from drained peatland forest soils for the greenhouse gas inventory of Finland

Abstract. In peatlands drained for forestry, the soil carbon (C) or carbon dioxide (CO2) balance is affected by both (i) higher heterotrophic CO2-C release from faster decomposing soil organic matter (SOM) and (ii) higher plant litter C input from more vigorously growing forests. This balance and other greenhouse gas (GHG) sinks and sources in managed lands are annually reported by national GHG inventories to the United Nations Climate Change Convention. In this paper, we present a revised, fully dynamic method for reporting the CO2 balance of drained peatland forest soils in Finland. Our method can follow temporal changes in tree biomass growth, tree harvesting and climatic parameters, and it is built on empirical regression models of SOM decomposition and litter input in drained peatland forests. All major components of aboveground and belowground litter input from ground vegetation as well as live trees and trees that died naturally are included, supplemented by newly acquired turnover rates of woody plant fine roots. Annual litter input from harvesting residues is calculated using national statistics of logging and energy use of trees. Leaching, which also exports dissolved C from drained peatlands, is not included. The results are reported as time series from 1990–2021 following the practice in the GHG inventory. Our revised method produces an increasing trend of annual emissions from 0.2 to 2.1 t CO2 ha−1 yr−1 for the period 1990–2021 in Finland (equal to a trend from 1.4 to 7.9 Mt CO2 yr−1 for the entire 4.3 Mha of drained peatland forests), with a statistically significant difference between the years 1990 and 2021. Across the period 1990–2021, annual emissions are on average 1.5 t CO2 ha−1 yr−1 (3.4 Mt CO2 yr−1 for 2.2 Mha area) in warmer southern Finland and −0.14 t CO2 ha−1 yr−1 (−0.3 Mt CO2 yr−1 for 2.1 Mha area) in cooler northern Finland. When combined with data on the CO2 sink created by the growing tree stock, in 2021 the drained peatland forest ecosystems were a source of 1.0 t CO2 ha−1 yr−1 (2.3 Mt CO2 yr−1) in southern Finland and a sink of 1.2 t CO2 ha−1 yr−1 (2.5 Mt CO2 yr−1) in northern Finland. We compare these results to those produced by the semi-dynamic method used earlier in the Finnish GHG inventory and discuss the strengths and vulnerabilities of the new revised method in comparison to more static emission factors.

The impact of vaccinations and chronic disease on COVID death rates

This study presents a theoretical and empirical regression model to measure the efficacy of vaccinations in reducing COVID death rates across states over the 3/10/21 to 12/28/22 period. During that period, it was estimated that the availability of vaccinations resulted in a reduction of 427,000 COVID deaths in the nation. To arrive at that estimate, other covariants were held constant. In particular, it was found that chronic disease should be included as an explanatory variable to arrive at unbiased measures of the efficacy of vaccinations in reducing deaths. In addition, the percentage of people over the age of 65 was found to be highly significant. The only ethnic/racial characteristic that was significant in explaining COVID deaths was the percentage of American Indians/Alaska Natives residing in a state. Other ethnic/racial characteristics, as well as variables representing population, density, governmental stringency, and income, were not significant over the period tested.

Machine-learning-based performance prediction of the energy pile heat pump system

The coefficient of performance (COP) is one of the important indicators for evaluating ground source heat pump systems, and a certain amount of machine learning algorithms have been applied for COP prediction. However, most of the current models use a large number of input variables, and it may be difficult to obtain all the required parameters for diagnosis of the sensors. Most of the influencing factors contain fluid temperature, which is not easy to measure, and there is a risk of leakage when sensors are deployed in the system. In this work, two-year field tests were conducted based on an energy pile heat pump system, where in situ results were used as sample points, and the more easily measured ambient temperature and humidity, room temperature and humidity, and hourly power consumption were used as input parameters. Spearman correlation analysis is used to rank the influencing factors, and the prediction of COP was developed by a hybrid artificial neural network (ANN), optimized via the particle swarm optimization algorithm (PSO), and finally compared with empirical models and their updates. Results showed that among the input variables, the most and least important factors affecting the COP are the ambient temperature and ambient humidity, respectively. The prediction results of PSO-ANN were in the best agreement with the measured values, and the accuracy was higher than that of the empirical regression models. The proposed method noted in this study can be used as a supplement to the current COP prediction.

Modelling the influence of climate change on characterization factors for copper terrestrial ecotoxicity

The use of copper-based preparations is a long-standing practice in viticulture to control vine fungal diseases, which has led to high copper concentrations in vineyard soils and impacts on off-target terrestrial organisms. Under projected climate change, some of the mechanisms and properties of soils that influence the extent of metals ecotoxicity impacts are projected to be altered, namely soil erosion, rainfall, temperature, and organic carbon content. In this context, and within the framework of life cycle assessment, this study aims to simulate the influence of projected changes in soil organic carbon, soil erosion, and rainfall on characterization factors (CFs) for copper terrestrial ecotoxicity across non-calcareous European vineyard soils. We employed empirical regression models to account for metal speciation and bioavailability as determined by soil characteristics. CFs were computed for a current scenario and mid-term future scenarios by 2050 determined across three Representative Concentration Pathways (RCP2.6, RCP4.5, and RCP8.5). Although future scenarios suggest that CFs may either increase or decrease, CFs are projected to increase for a larger share of the European vineyard surface, which may lead to higher terrestrial ecotoxicity impacts. The RCP4.5 scenario projects the highest increases in CFs, with a 27% rise in the median CF in comparison with the current scenario. Whereas the RCP8.5 and RCP2.6 scenarios project a 19% and 14% increase in median CFs, respectively. The changes in CFs were determined principally by the changes in copper bioavailability driven by projected changes in soil organic carbon. However, the spatial variability of CFs was larger than the temporal variation, with a variation of roughly 2 orders of magnitude across the analyzed scenarios. This study highlights the relevance of integrating spatial differentiation and the influence of projected climate change in the characterization modelling of copper terrestrial ecotoxicity.

Prediction accuracies of ray-tracing and regression models in open-plan offices

The purpose was to determine the prediction accuracies of an empirical regression model (Model 1) and a ray-tracing model (Model 2) in acoustically different open-plan offices. The measurement data was obtained from an open-plan office (81 m2) in which 22 completely different room acoustic conditions were built using alternative room absorption areas in ceiling, walls, and screens, and alternative screen heights. The sound pressure level (SPL) produced by an omnidirectional sound source was measured in six positions in each condition. The corresponding SPLs in 22 conditions were predicted using Models 1 and 2. The prediction accuracies were determined for the A-weighted SPL of speech in each position. In addition, the prediction accuracy was determined for the basic quantities of ISO 3382 standard: spatial decay rate of speech, D2,S, A-weighted SPL of speech at 4 m distance, Lp,A,S,4m, comfort distance, rC, and mean reverberation time T. The mean prediction accuracy of A-weighted SPL of speech was 4.2 dB for Model 1 and 3.4 dB for Model 2. Model 1 predicted D2,S, rC, and T more accurately than Model 2, while the opposite ranking order was found for Lp,A,S,4m. The prediction accuracies of both Models were sufficient when the effects of sound absorption materials and screen heights is simulated. The finding suggests that simple empirical regression models can provide sufficient prediction accuracy in most room acoustic conditions of open-plan offices.

Nitrogen leaching and groundwater N contamination risk in saffron/wheat intercropping under different irrigation and soil fertilizers regimes

The ever-rising trend of nitrate leaching from the agricultural production systems is a major risk to the contamination of ground- and surface-waters and should be addressed. But so far, there has been no study on the reduction of nitrate leaching from saffron fields through intercropping. Saffron growers can make a sustainable use of the saffron inter-row spaces through the strategy of winter-wheat/saffron base intercropping system to reduce nitrate leaching. During four years of study, in a set of lysimeters, effects of two cropping systems (saffron mono-cropping and saffron–wheat intercropping), application of two sources of nitrogen (organic cow manure and chemical granular urea) and four irrigation regimes [40, 60, 80, and 100% of the standard crop evapotranspiration (ETc)] on plant nitrogen and phosphorus uptake, nitrogen leaching and nitrogen and phosphorus efficiencies were investigated. The optimum irrigation regime was experienced at 60% ETc (with irrigation application efficiency of 60%, equivalent to 100%ETc) where the highest saffron and wheat nutrient (nitrogen and phosphorus) uptake, nutrient (nitrogen and phosphorus) harvest indices, nutrient acquisition and use efficiencies, corm, saffron, and grain yields and lowest nitrogen loss was achieved. Moreover, manure application indicated 12, 42, 50 and 46% lower amounts of drained water, leachate nitrate nitrogen concentration, total leached nitrogen and N losses (other than N leaching), respectively, in comparison to the urea source of nitrogen showing the lower risk of groundwater nitrate pollution. Manure application showed 9, 8 and 9% increase in the concentration of corm nitrogen, phosphorus and protein, respectively, in comparison to urea application treatment. Saffron corm and stigma yields, irrigation and economic water productivities, corm nitrogen use efficiency and saffron-plant-nitrogen-acquisition efficiency in manure application surpassed respectively by 21, 25, 20, 17, 39 and 49% compared with the chemical source of nitrogen. Intercropping showed 10, 11, 23 and 64% lower amounts of drained water, nitrate concentration in drainage water, seasonal leached nitrate and N losses (other than N leaching), respectively compared to saffron sole cropping which reduces the risk of groundwaters nitrate contamination. For all the experimental treatments, empirical regression models were derived for estimation of seasonal leached nitrate based on the seasonal drained water. Intercropping saffron with winter wheat, application of organic cow manure and adopting irrigation regime of 60% ETc is an innovative system of saffron production which mitigates the risk of groundwater nitrate contamination and increases irrigation and economic water productivities. Saffron growers can make sustainable and clean use of the inter-row spaces of the saffron crop to grow winter wheat in order to obtain higher economic water productivity and lower groundwater nitrate pollution, and it is highly recommended to maintain a sustainable environment.

Stochastic simulation of velocity pulses of near-fault ground motions based on multivariate copula modeling

Due to the featured high-amplitude, long-period pulses in velocity or displacement time–history, the near-fault ground motion may have a stronger destructive effect on engineering structures. Therefore, studying velocity pulses modeling of near-fault ground motions has been an important topic in the earthquake engineering community. However, the variability and probabilistic correlation of pulse model parameters have not received enough attention. To this end, an elegant method for stochastic simulation of velocity pulses of near-fault ground motions is proposed based on the classical Gabor wavelet modeling and probabilistic correlation analysis of associated model parameters by multivariate copula modeling. Especially, an efficient scheme of iteratively determining the optimal copulas pertinent to edges of a complex vine structure is developed, and two parameters relevant to the near-fault ground motion records, i.e., the rupture distance and the strongest pulse orientation, are introduced which provides a physical basis for addressing the dependence structure among variables. For illustrative purposes, the ground accelerograms recorded by near-fault instruments in the 1999 Chi-Chi earthquake are used. It is shown that the multivariate copula modeling enables the efficient representation of probabilistic correlation between model parameters which is consistent with their physical mechanism. The time histories and spectral characteristics of velocity pulses of near-fault ground motions and their stochastic fluctuations with respect to the rupture distance are well simulated. While the empirical linear regression model does not meet the requirements of random vibration analysis and thus cannot guide the safety design of seismic structures.

New emission inventory reveals termination of global dioxin declining trend.

Accurate estimates of spatiotemporally resolved Polychlorinated dibenzo-p-dioxins (PCDD/Fs, or dioxins) emissions are critical for understanding their environmental fate and associated health risks. In this study, by utilizing an empirical regression model for PCDD/Fs emissions, we developed a global emission inventory for 17 toxic PCDD/Fs congeners from 8 source sectors with a spatial resolution of 1° ×1° from 2002 to 2018. The results show that PCDD/Fs emissions decreased by 25.7 % (12.5kg TEQ) between 2002 and 2018, mostly occurring in upper- and lower-middle income countries. Globally, open-burning processes, waste incineration, ferrous and nonferrous metal production sectors and heat and power generation were the major source sectors of PCDD/Fs. Spatially, high PCDD/Fs emissions were mainly identified in East and South Asia, Southeast Asia, and part of Sub-Saharan Africa. We find that the declining trend of dioxin emissions over the past decades terminated from the early 2010s due to increasing significance of wildfire induced emissions in the total emission. The PCDD/Fs emission inventory developed in the present study was verified by inputting the inventory as initial conditions into an atmospheric transport model, the Canadian Model for Environmental Transport of Organochlorine Pesticides (CanMETOP), to simulate PCDD/Fs concentrations in air and soil. The predicted concentrations were compared to field sampling data. The good agreement between the modeled and measured concentrations demonstrates the reliability of the inventory.

Estimation and mapping of water quality parameters using satellite images: a case study of Two Rivers Dam, Kenya

AbstractThe continuous water quality monitoring (WQM) of watersheds and the existing water supplies is a crucial step in realizing sustainable water development and management. However, the conventional approaches are time-consuming, labor intensive, and do not give spatial–temporal variations of the water quality indices. The advancements in remote sensing techniques have enabled WQM over larger temporal and spatial scales. This study used satellite images and an empirical multivariate regression model (EMRM) to estimate chlorophyll-a (Chl-a), total suspended solids (TSS), and turbidity. Furthermore, ordinary Kriging was applied to generate spatial maps showing the distribution of water quality parameters (WQPs). For all the samples, turbidity was estimated with an R2 and Pearson correlation coefficient (r) of 0.763 and 0.818, respectively while TSS estimation gave respective R2 and r values of 0.809 and 0.721. Chl-a was estimated with accuracies of R2 and r of 0.803 and 0.731, respectively. Based on the results, this study concluded that WQPs provide a spatial–temporal view of the water quality in time and space that can be retrieved from satellite data products with reasonable accuracy.

Experimental Investigation on the Wear Performance of Nano-Additives on Degraded Gear Lubricant

This study investigates the degradation of a commercially available gear lubricant and the potential of nano-additives to mitigate such degradation. Initially, we performed an experimental study on the chemical degradation of commercially available API GL-4 EP90 gear lubricant by mixing the different concentrations of aqueous hydrochloric acid (aqueous HCl) varying from 0.0005% v/v up to 0.0025% v/v, while maintaining overall water content in the oil below the prescribed limits. The degradation was monitored using the pH value, total acid number (TAN) value, and attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) analysis. The experiments were performed on a developed gear test rig using conventional gear oil as well as chemically aged gear oil, and the corresponding results of pH value and wear debris were recorded. Based on the results, an empirical regression model between the concentration of aqueous HCl and lubricant aging time has been established. Under chemically aged lubricant, severe wear of gear was observed, which motivated us to explore suitable nano-additive to minimize the gear wear. Initially, three nano-additives—graphite, graphene, and “graphene oxide functionalized with silicon oxide (GO@SiO2)”—were chosen. A series of tests were conducted using the design of experiments method (L8 and L16 orthogonal array) to investigate the effect of nano-additives and to find the optimum additive for wear performance. Finally, experiments were conducted on gear setup using the degraded lubricant optimized with nano-additive. Overall results indicate a very significant contribution of nano-additives in decreasing gear wear.

The PoliTO–UniRoma1 database of cyclic and dynamic laboratory tests: assessment of empirical predictive models

The soil nonlinear hysteretic behaviour is usually described, in the moderate strain range, through the shear modulus reduction and material damping ratio (MRD) curves. In common practice, in absence of specific laboratory tests, the curves are estimated by employing empirical regression models. Such predictive models, typically calibrated on large experimental datasets, correlate the soil response to its physical properties. This research fits within this context, presenting a comprehensive database of cyclic and dynamic laboratory tests conducted on natural Italian soils. The database, publicly available as supplementary data of the paper, contains the results of the tests conducted by the geotechnical laboratories of the Politecnico di Torino (Turin, Italy) and the Sapienza Università di Roma (Rome, Italy) over the past 30 years. The experimental data are employed to assess the performance of some widely used empirical models in predicting the MRD curves of natural uncemented fine-grained soils, emphasizing the importance of using an independent dataset for conducting a reliable statistical analysis. The results show that the use of many soil parameters as proxies for predicting the soil response does not necessarily lead to an improvement in the performance of the model. Therefore, according to Occam’s razor principle, simple models are to be preferred.

Estimating local-scale forest GPP in Northern Europe using Sentinel-2: Model comparisons with LUE, APAR, the plant phenology index, and a light response function

Northern forest ecosystems make up an important part of the global carbon cycle. Hence, monitoring local-scale gross primary production (GPP) of northern forest is essential for understanding climatic change impacts on terrestrial carbon sequestration and for assessing and planning management practices. Here we evaluate and compare four methods for estimating GPP using Sentinel-2 data in order to improve current available GPP estimates: four empirical regression models based on either the 2-band Enhanced Vegetation Index (EVI2) or the plant phenology index (PPI), an asymptotic light response function (LRF) model, and a light-use efficiency (LUE) model using the MOD17 algorithm. These approaches were based on remote sensing vegetation indices, air temperature (Tair), vapor pressure deficit (VPD), and photosynthetically active radiation (PAR). The models were parametrized and evaluated using in-situ data from eleven forest sites in North Europe, covering two common forest types, evergreen needleleaf forest and deciduous broadleaf forest. Most of the models gave good agreement with eddy covariance-derived GPP. The VI-based regression models performed well in evergreen needleleaf forest (R2 = 0.69–0.78, RMSE = 1.97–2.28 g C m−2 d−1, and NRMSE = 9–11.0%, eight sites), whereas the LRF and MOD17 performed slightly worse (R2 = 0.65 and 0.57, RMSE = 2.49 and 2.72 g C m−2 d−1, NRMSE = 12 and 13.0%, respectively). In deciduous broadleaf forest all models, except the LRF, showed close agreements with the observed GPP (R2 = 0.75–0.80, RMSE = 2.23–2.46 g C m−2 d−1, NRMSE = 11–12%, three sites). For the LRF model, R2 = 0.57, RMSE = 3.21 g C m−2 d−1, NRMSE = 16%. The results highlighted the necessity of improved models in evergreen needleleaf forest where the LUE approach gave poorer results., The simplest regression model using only PPI performed well beside more complex models, suggesting PPI to be a process indicator directly linked with GPP. All models were able to capture the seasonal dynamics of GPP well, but underestimation of the growing season peaks were a common issue. The LRF was the only model tending to overestimate GPP. Estimation of interannual variability in cumulative GPP was less accurate than the single-year models and will need further development. In general, all models performed well on local scale and demonstrated their feasibility for upscaling GPP in northern forest ecosystems using Sentinel-2 data.

Modeling hydraulic fracture fluid efficiency in tight gas reservoirs using non-linear regression and a back-propagation neural network

ABSTRACT This study introduces a back-propagation (BP) neural network model and a regression model for estimating the fracture fluid efficiency based on a data set consisting of 1261 staged and ramped simulation runs of tight gas reservoirs subjected to hydraulic fracturing treatment. Data were generated using a 3-D commercial simulator which is a versatile software portfolio that models many well configurations, proppant placement, and fracture geometries. The BP network inputs consist of shear rate/fracture conductivity ratio, the injection rate, reservoir permeability, formation closure stress, reservoir thickness, effective viscosity, and fracture height. The neural network model was able to generate satisfactory estimates of the fracture fluid efficiency for the training dataset, and for the blind testing data. An average error of approximately 2.5% was obtained for the training set, and an average error of 3% was obtained for the testing set. An empirical non-linear regression model has been constructed based on dimensionless groups derived by applying dimensional analysis to a set of variables consisting of the maximum fracture width, fracture length, fracture height, effective viscosity, shear rate/fracture conductivity ratio, reservoir thickness, injection rate, reservoir permeability, and formation closure stress. The average error for estimating the fluid efficiency using the non-linear regression empirical model was approximately 6.17%. Since the non-linear regression model has an explicit formulation, it is easier to apply than the neural network model. The empirical regression model estimates of the fluid efficiency appeared to be unbiased and were more precise than those estimates obtained using either the KGD or the PKN 2-D models. The introduced BP model and the non-linear regression model offer fast and inexpensive alternatives to the application of three-dimensional simulators for estimating the fluid efficiency.

Estimation of sound speed profiles based on remote sensing parameters using a scalable end-to-end tree boosting model

IntroductionIn underwater acoustic applications, the three-dimensional sound speed distribution has a significant impact on signal propagation. However, the traditional sound speed profile (SSP) measurement method requires a lot of manpower and time, and it is difficult to popularize. Satellite remote sensing can collect information on a large ocean surface area, from which the underwater information can be derived.MethodIn this paper, we propose a method for reconstructing the SSP based on an extensible end-to-end tree boosting (XGBoost) model. Combined with satellite remote sensing data and Argo profile data, it extracts the characteristic matrix of the SSP and analyzes the contribution rate of each order matrix to reduce the introduction of noise. The model inverts the SSP above 1000 m in the South China Sea by using the root mean square error (RMSE) as the precision evaluation index.ResultThe results showed that the XGBoost model could better reconstruct the SSP above 1000 m, with a RMSE of 1.75 m/s. Compared with the single empirical orthogonal function regression (sEOF-r) model of the linear regression method, the RMSE of the XGBoost model was reduced by 0.59 m/s.DiscussionFor this model, the RMSE of the inversion results was smaller, the robustness was better, and the regression performance was superior to that of the sEOF-r model at different depths. This study provided an efficient tree boosting model for SSP reconstruction, which could reliably and instantaneously monitor the 3D sound speed distribution.

Mapping snow density through thermal inertia observations

Snow, as a fundamental reservoir of freshwater, is a crucial natural resource. Specifically, knowledge of snow density spatial and temporal variability could improve modelling of snow water equivalent, which is relevant for managing freshwater resources in context of ongoing climate change. The possibility of estimating snow density from remote sensing has great potential, considering the availability of satellite data and their ability to generate efficient monitoring systems from space.In this study, we present an innovative method that combines meteorological parameters, satellite data and field snow measurements to estimate thermal inertia of snow and snow density at a catchment scale. Thermal inertia represents the responsiveness of a material to variations in temperature and depends on the thermal conductivity, density and specific heat of the medium. By exploiting Landsat 8 data and meteorological modelling, we generated multitemporal thermal inertia maps in mountainous catchments in the Western European Alps (Aosta Valley, Italy), from incoming shortwave radiation, surface temperature and snow albedo. Thermal inertia was then used to develop an empirical regression model to infer snow density, demonstrating the possibility of mapping snow density from optical and thermal observations from space. The model allows for estimation of snow density with R2CV and RMSECV of 0.59 and 82 kg m−3, respectively. Thermal inertia and snow density maps are presented in terms of the evolution of snow cover throughout the hydrological season and in terms of their spatial variability in complex topography. This study could be considered a first attempt at using thermal inertia toward improved monitoring of the cryosphere. Limitations of and improvements to the proposed methods are also discussed.This study may also help in defining the scientific requirements for new satellite missions targeting the cryosphere. We believe that a new class of Earth Observation missions with the ability to observe the Earth's surface at high spatial and temporal resolution, with both day and night-time overpasses in both optical and thermal domain, would be beneficial for the monitoring of seasonal snowpacks around the globe.

CROP YIELD ESTIMATION IN THE NORTH CHINA PLAIN FROM 2001 TO 2016 USING MULTI-SOURCE REMOTE SENSING DATA AND PROCESS-BASED FGM MODEL

Abstract. Gross primary productivity (GPP) is an essential indicator of vegetation growth that reflects ecosystem function. GPP is the original source of energy entering cropland ecosystem and thus could serve as a direct indicator of crop yield. In the context of increasing population, changing climate, and decreasing available resources, accurate monitoring and forecasting of food and crop yields play an essential role in sustainable human development. In this study, the process-based Farquhar GPP model (FGM) driven by multisource remote sensing data was implemented to estimate the spatial and temporal dynamics of GPP in crop-growing areas of the North China Plain from 2001 to 2016. We found that the GPP of crops in the North China Plain is relatively high in the southern provinces while lower in the northern part. The GPP values showed a significant increasing trend from 2001 to 2016 (+2.19 Mt C yr−1, P<0.05). Based on crop yield statistical yearbook, we found that GPP is well correlated with crop yield (R2 = 0.98, RMSE = 10.4 Mt yr−1). Thus, we constructed an empirical regression model between GPP and crop yield (i.e., ‘GPP-yield’ empirical model). Finally, time-series GPP data and the ‘ GPP-yield ’ model were applied the crop yield in the North China Plain with spatial and temporal continuity. We found that the crop yield in the North China Plain changed in accordance with GPP, and also showed a significant increasing trend from 2001 to 2016, with a mean increasing rate of +2.84 Mt yr−1 (P<0.05, R2 = 0.16, RMSE = 31.73 Mt yr−1). This study proved an example of large-scale crop yield estimation using multi-source remote sensing data.

Estimating LAI for Cotton Using Multisource UAV Data and a Modified Universal Model

Leaf area index(LAI) is an important indicator of crop growth and water status. With the continuous development of precision agriculture, estimating LAI using an unmanned aerial vehicle (UAV) remote sensing has received extensive attention due to its low cost, high throughput and accuracy. In this study, multispectral and light detection and ranging (LiDAR) sensors carried by a UAV were used to obtain multisource data of a cotton field. The method to accurately relate ground measured data with UAV data was built using empirical statistical regression models and machine learning algorithm models (RFR, SVR and ANN). In addition to the traditional spectral parameters, it is also feasible to estimate LAI using UAVs with LiDAR to obtain structural parameters. Machine learning models, especially the RFR model (R2 = 0.950, RMSE = 0.332), can estimate cotton LAI more accurately than empirical statistical regression models. Different plots and years of cotton datasets were used to test the model robustness and generality; although the accuracy of the machine learning model decreased overall, the estimation accuracy based on structural and multisources was still acceptable. However, selecting appropriate input parameters for different canopy opening and closing statuses can alleviate the degradation of accuracy, where input parameters select multisource parameters before canopy closure while structural parameters are selected after canopy closure. Finally, we propose a gap fraction model based on a LAImax threshold at various periods of cotton growth that can estimate cotton LAI with high accuracy, particularly when the calculation grid is 20 cm (R2 = 0.952, NRMSE = 12.6%). This method does not require much data modeling and has strong universality. It can be widely used in cotton LAI prediction in a variety of environments.

A Semi Empirical Regression Model for Critical Dent Depth of Externally Corroded X65 Gas Pipeline.

External corrosion dent is a common type of compound dent. On the one hand, this type of compound dent reduces the bearing area and bearing capacity of the pipeline. On the other hand, it leads to an increase in the stress–strain concentration in the dent and reduces the anti-fatigue load capacity of the pipeline, which is more harmful to the service safety of the pipeline than the simple dent. In this study, the reliability of the modeling method was verified by the numerical inversion of the full-size dented pipe test. A three-dimensional finite element model for a pipe with a small corrosion dent was established by analyzing the internal detection data on corrosion defects of pipes with a diameter of 813 mm. The failure criterion of the corrosion dent pipe and the calculation method of the critical dent depth were determined. The influence of corrosion depth, length, width, internal pressure load, curvature radius of indenter, and diameter–thickness pipeline ratio on critical dent depth was investigated. Finally, a critical dent-depth prediction formula was developed based on the numerical results. This study provides a reference and significant guidance for the applicability evaluation of corroded sunken pipelines.

Estimation of Sound Speed Profiles Using a Random Forest Model with Satellite Surface Observations

Sound speed profile (SSP) inversion is usually performed by linear statistical regression, such as the single empirical orthogonal function regression (sEOF-r) model. However, due to the complex dynamic activities of the ocean, the relationship between parameters is not strictly linear, often resulting in an unsatisfactory inversion result. In this study, an algorithm based on the random forest (RF) integrated learning model, for SSP inversion, was proposed. Using the sea surface temperature anomaly (SSTA) and sea surface height anomaly (SSHA) data, the sound speed profile of the upper 1000 m layer in the South China Sea was reconstructed, and its accuracy was evaluated through the root mean square error (RMSE). The accuracy of the evaluation demonstrated that the RF model proposed here could reconstruct the SSP in the upper 1000 m layer better than the sEOF-r can. Compared with the latter, the average reconstruction accuracy of the RF model was improved by 0.56 m/s. The linear regression of the sEOF-r model fell short of expectations in the regression between surface and subsurface parameters. By removing the constraints of linear inversion, the nonlinear regression of the RF model showed a smaller RMSE and better robustness in the reconstruction process and was superior to the sEOF-r model at all depths. As a result, it provided an effective integrated learning model for SSP reconstruction.

A fuzzy empirical quantile-based regression model based on triangular fuzzy numbers

A fuzzy empirical quantile-based regression model based on triangular fuzzy numbers