Mean Absolute Bias Research Articles

EQA: There's not a glitch in the matrix. Investigation of CRP bias on the Roche Cobas c701.

The Roche Cobas c Tina-quant C-Reactive Protein (CRP) IV method has shown persistent, significant negative bias compared to other methodologies in the UK NEQAS external quality assurance (EQA) scheme. The aim of this study was to compare CRP concentrations determined by this restandardized 4th generation assay with another commercially available method to assess the extent of any bias and impact on clinical cut-offs. CRP concentrations in 38 anonymized patient serum samples were determined by a particle-enhanced immunoturbidimetric assay on the Roche Cobas c701 module and a latex assay on the Beckman Coulter AU5800 analyser over a concentration range of 1-308mg/L. 97.4% samples analysed by the Roche Tina-quant CRP IV method demonstrated a negative bias compared to the Beckman AU latex method. The mean absolute bias was -4.12mg/L (range: -19.13-10.93mg/L; 95% CI: -17.25-9.01mg/L). The mean relative difference was -15.5% (range: -40-4%; 95% CI: -33.03-1.94%). This bias was seen across the range of samples assayed, including at clinically significant cut-offs of 5mg/L (-24% bias), 20mg/L (-5%) and 100mg/L (-13%). The negative bias of the Roche method demonstrated in the EQA scheme appears to reflect genuine differences in patient results, rather than an EQA matrix effect, despite re-standardization of the CRP assay. This report indicates that clinical cut-offs and reference ranges may be method-dependent, which should be reflected in published guidance and interpretation provided by laboratories.

Fault Damage Zone Effects on Ground Motions during the 2019 Mw 7.1 Ridgecrest, California, Earthquake

ABSTRACT We have simulated 0–3 Hz deterministic wave propagation in the Southern California Earthquake Center Community Velocity Model (CVM) version CVM-S4.26-M01 for the 2019 Mw 7.1 Ridgecrest earthquake. A data-constrained high-resolution fault zone model (Zhou et al., 2022) is incorporated into the CVM to investigate the effects of the near-fault low-velocity zone (LVZ) on the resulting ground motions, constrained by strong-motion data recorded at 161 stations. The finite-fault source used for the simulation of the Ridgecrest event was obtained from the Liu et al. (2019) kinematic inversion, enriched by noise following a von Karman correlation function above ∼1 Hz with a f−2 high-frequency decay. Our results show that the heterogeneous near-fault LVZ inherent to the fault zone structure significantly perturbs the predicted wave field in the near-source region, in particular by more accurately generating Love waves at its boundaries. The fault zone decreases the 0.1–0.5 Hz mean absolute Fourier amplitude spectrum bias to seismic recordings for all sites in the model and in the Los Angeles basin area (∼200 km from the source) by 16% and 26%, respectively. The fault zone structure generally improves modeling of the long-period features in the data and lengthens the coda-wave trains, in better agreement with observations. The favorable fit to data was obtained with a model including high-resolution surface topography, a 700-m-thick geotechnical layer and frequency-dependent anelastic attenuation in the model domain, with QS=0.1VS and QS(f)=0.1VSf0.5 (VS in m/s) for frequencies lower and higher than 1 Hz, respectively. We recommend that a data-constrained fault zone velocity structure, where available, be included in ground-motion modeling to obtain the least-biased fit to observed seismic data.

Multi-energy CT material decomposition using Bayesian deep convolutional neural network with explicit penalty of uncertainty and bias.

Convolutional neural network (CNN)-based material decomposition has the potential to improve image quality (visual appearance) and quantitative accuracy of material maps. Most methods use deterministic CNNs with mean-square-error loss to provide point-estimates of mass densities. Point estimates can be over-confident as the reliability of CNNs is frequently compromised by bias and two major uncertainties - data and model uncertainties originating from noise in inputs and train-test data dissimilarity, respectively. Also, mean-square-error lacks explicit control of uncertainty and bias. To tackle these problems, a Bayesian dual-task CNN (BDT-CNN) with explicit penalization of uncertainty and bias was developed. It is a probabilistic CNN that concurrently conducts material classification and quantification and allows for pixel-wise modeling of bias, data uncertainty, and model uncertainty. CNN was trained with images of physical and simulated tissue-mimicking inserts at varying mass densities. Hydroxyapatite (nominal density 400mg/cc) and blood (nominal density 1095mg/cc) inserts were placed in different-sized body phantoms (30 - 45cm) and used to evaluate mean-absolute-bias (MAB) in predicted mass densities across different images at routine- and half-routine-dose. Patient CT exams were collected to assess generalizability of BDT-CNN in the presence of anatomical background. Noise insertion was used to simulate patient exams at half- and quarter-routine-dose. The deterministic dual-task CNN was used as baseline. In phantoms, BDT-CNN improved consistency of insert delineation, especially edges, and reduced overall bias (average MAB for hydroxyapatite: BDT-CNN 5.4mgHA/cc, baseline 11.0mgHA/cc and blood: BDT-CNN 8.9mgBlood/cc, baseline 14.0mgBlood/cc). In patient images, BDT-CNN improved detail preservation, lesion conspicuity, and structural consistency across different dose levels.

Development criterion of estimating hourly global solar radiation for all sky conditions in China

The high quality hourly global solar radiation data becomes more and more important under the trend of low building energy consumption and carbon emission. The models based on daily global solar radiation and hourly meteorological parameters are two kinds of efficient methods to compensate for measured data deficiency. However, the applicable climate and sky conditions of two models are needed to be clarified. Based on 45 stations in China, the coupling correlations between hourly global solar radiation and meteorological parameters are obtained by correlation analysis. The eight decomposition models and four models based on the main hourly meteorological parameters are developed and evaluated for climates and sky conditions using the mean absolute bias error, relative root mean squared error, and Nash-Sutcliffe efficiency coefficient. Considering the advantages of two kinds of models, the criterion of estimating hourly global solar radiation data is developed. The results show that hourly sunshine duration and hourly surface temperature are the main factors reflecting the change of hourly global solar radiation. In humid and sub-humid climates, the hourly relative humidity can further characterize the fluctuation of hourly global solar radiation. CPR model performs best among decomposition models, especially in sunny sky conditions. The meteorological parameters models own higher accuracy in cloudy conditions. In regions with hourly sunshine duration measurement, CPR model is recommended when daily clearness index is higher than 0.7 and lower than 0.2. When daily clearness index varies from 0.2 to 0.7, the model should be established by selecting hourly sunshine duration and hourly surface temperature in arid and semi-arid climates and hourly sunshine duration, hourly surface temperature and hourly relative humidity in humid and subhumid climates, respectively. In regions without hourly sunshine duration measurement, CPR model should be used regardless of weather conditions and climate types.

Deep learning mapping of surface MDA8 ozone: The impact of predictor variables on ozone levels over the contiguous United States

The limited number of ozone monitoring stations imposes uncertainty in various applications, calling for accurate approaches to capturing ozone values in all regions, particularly those with no in-situ measurements. This study uses deep learning (DL) to accurately estimate daily maximum 8-hr average (MDA8) ozone and examines the spatial contribution of several factors on ozone levels over the contiguous U.S. (CONUS) in 2019. A comparison between in-situ observations and DL-estimated MDA8 ozone values shows a correlation coefficient (R) of 0.95, an index of agreement (IOA) of 0.97, and a mean absolute bias (MAB) of 2.79 ppb, highlighting the promising performance of the deep convolutional neural network (Deep-CNN) at estimating surface MDA8 ozone. Spatial cross-validation also confirms the high spatial accuracy of the model, which obtains an R of 0.91, and IOA of 0.96 and an MAB of 3.46 ppb when it is trained and tested on separate stations. To interpret the black-box nature of our DL model, we use Shapley additive explanations (SHAP) to generate a spatial feature contribution map (SFCM), the results of which confirm an advanced ability of Deep-CNN to capture the interactions between most predictor variables and ozone. For instance, the model shows that solar radiation (SRad) SFCM, with higher values, enhances the formation of ozone, particularly in the south and southwestern CONUS. As SRad triggers ozone precursors to produce ozone via photochemical reactions, it increases ozone concentrations. The model also shows that humidity, with its low values, increases ozone concentrations in the western mountainous regions. The negative correlation between humidity and ozone levels can be attributed to factors such as higher ozone decomposition resulting from increased levels of humidity and OH radicals. This study is the first to introduce the SFCM to investigate the spatial role of predictor variables on changes in estimated MDA8 ozone levels.

Global Navigation Satellite System-Based Retrieval of Precipitable Water Vapor and Its Relationship with Rainfall and Drought in Qinghai, China

Qinghai Province is situated deep in inland China, on the Qinghai-Tibet plateau, and it has unique climate change characteristics. Therefore, understanding the temporal and spatial distributions of water vapor in this region can be of great significance. The present study applied global navigation satellite system (GNSS) technology to retrieve precipitable water vapor (PWV) in Qinghai and analyzed its relationship with rainfall and drought. Firstly, radiosonde (RS) data is used to verify the precision of the surface pressure (P) and temperature (T) from the fifth-generation atmosphere reanalysis data set (ERA5) of the European Centre for Medium-Range Weather Forecasts (ECMWF), as well as the zenith troposphere delay (ZTD), calculated based on the data from continuously operating reference stations (CORS) in Qinghai. Secondly, a regional atmospheric weighted mean temperature (Tm) (QH-Tm) model was developed for Qinghai based on P, T, and relative humidity, as well as the consideration of the influence of seasonal changes in Tm. Finally, the PWV of each CORS in Qinghai was calculated using the GNSS-derived ZTD and ERA5-derived meteorological data, and its relationship with rainfall and drought was evaluated. The results show that the ERA5-derived P and T have high precision, and their average root mean square (RMS), mean absolute error (MAE) and bias were 1.06/0.85/0.01 hPa and 2.98/2.42/0.03 K, respectively. The RMS, MAE and bias of GNSS-derived ZTD were 13.2 mm, 10.3 mm and −1.8 mm, respectively. The theoretical error for PWV was 1.98 mm; compared with that of RS- and ERA5-derived PWV, the actual error was 2.69 mm and 2.16 mm, respectively. In addition, the changing trend of GNSS-derived PWV was consistent with that of rainfall events, and it closely and negatively correlated with the standardized precipitation evapotranspiration index. Therefore, the PWV retrieved from GNSS data in this study offers high precision and good feasibility for practical applications; thus, it can serve as a crucial tool for investigating water vapor distribution and climate change in Qinghai.

Analysis of systematic biases in tropospheric hydrostatic delay models and construction of a correction model

Abstract. In the field of space geodetic techniques, such as global navigation satellite systems (GNSSs), tropospheric zenith hydrostatic delay (ZHD) is chosen as the a priori value of tropospheric total delay. Therefore, the inaccuracy of ZHD will definitely affect parameters like the wet delay and the horizontal gradient of tropospheric delay, accompanied by an indirect influence on the accuracy of geodetic parameters, if not dealt with well at low elevation angles. In fact, however, the most widely used ZHD model currently seems to contain millimeter-level biases from the precise integral method. We explored the bias of traditional ZHD models and analyzed the characteristics in different aspects on a global annual scale. It was found that biases differ significantly with season and geographical location, and the difference between the maximum and minimum values exceeds 30 mm, which should be fully considered in the field of high-precision measurement. Then, we constructed a global grid correction model, which is named ZHD_crct, based on the meteorological data of the year 2020 from the ECMWF (European Centre for Medium-Range Weather Forecasts), and it turned out that the bias of traditional models in the current year could be reduced by ∼ 50 % when the ZHD_crct was added. When we verified the effect of ZHD_crct on the biases in the next year, it worked almost the same as the former year. The mean absolute biases (MABs) of ZHD will be narrowed within ∼ 0.5 mm for most regions, and the SDs (standard deviations) will be within ∼ 0.7 mm. This improvement will be helpful for research on meteorological phenomena as well.

Analytical validation of the modified Westergren method on the automated erythrocyte sedimentation rate analyzer CUBE 30 touch.

Analytical validation of automated erythrocyte sedimentation rate (ESR) analyzers is necessary prior to their implementation into routine practice. Our aim was toperform the analytical validation of the modified Westergren method applied on the CUBE 30 touch analyzer (Diesse, Siena, Italy). Validation included determination of within-run and between-run precision following the Clinical and Laboratory Standards Institute EP15-A3 protocol, comparison with the reference Westergren method, sample stability assessment at both room temperature and 4°C, after 4, 8 and24-h storage, and checking the extent of hemolysis and lipemia interference. Coefficients of variation (CVs) for within-run precision were 5.2% for the normal and 2.6% for the abnormal range, while between-run CVs were 9.4 and 2.2%, respectively. Comparison with the Westergren method (n=191) yielded Spearman's correlation coefficient of 0.93, noconstant nor proportional difference [y=0.4 (95% CI:-1.7-1.0)+1.06 (95% CI: 1.00-1.14)x] and a non-significant mean absolute bias of-2.6mm (95% CI:-5.3-0.2). Lower comparability was evidenced with increasing ESR values, with both constant and proportional differences for ESR values between 40 and 80mm, and above 80mm. Sample stability was not compromised up to 8-h storage both at room temperature (p=0.054) and 4°C (p=0.421). Hemolysis did not affect ESR measurement up to 1.0g/L of free hemoglobin (p=0.089), while lipemia index above 5.0g/L affects the ESR result (p=0.004). This study proved that CUBE 30 touch provides reliable ESR measurement and satisfactory comparability with the reference Westergren methods, with minor variation related to methodological differences.

Aboveground Biomass Models for Indigenous Tree Species in the Dry Afromontane Forest, Central Ethiopia

The application of biomass models for quantifying forests’ above-ground biomass is essential for sustainable forest management. However, lack of knowledge in modelig biomass of individual tree growth hinders the sustainable management of Dry Afromontane forests. In this study, models to estimate above-ground biomass were developed for Rhus ruspolii, Ekebergia capensis, and Nuxia congesta. To develop the models, a total of 45 trees from different diameter classes were selected, felled, and divided into different biomass compartments. For the model’s development, diameter at breast height (DBH), total height (TH), diameter at stump height (DSH), and wood density (WD) were used as independent variables. Models’ performances were evaluated using RSE, adjusted coefficient of determination, and AIC. Also, model validations were done by using rRMSE, mean absolute deviation, bias, and coefficient of variation. Models 5 (Adj-R2 = 0.92), 6 (Adj-R2 = 0.97), and 8 (Adj-R22 = 0.82) were the best fitted models for Nuxia congesta, Ekebergia capensis, and Rhus ruspolii, respectively. The average wood densities of Ekebergia capensis, Nuxia congesta, and Rhus ruspolii were 0.59, 0.50, and 0.69, respectively. The variation between observed biomass and estimated biomass using new models was statistically not significant (p>0.05). Thus, the biomass models developed here can be important tools to accurately estimate above-ground biomass in the Menagesha Suba forest and can be integrated into decision support tools.

Comparison between the ASTER and ECOSTRESS global emissivity datasets

High spatial resolution global land surface emissivity datasets are valuable for various applications. The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global Emissivity Dataset (GED) is the only accessible global emissivity data with a high spatial resolution (∼100 m). Since the launch of the ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS) in June 2018, another high spatial resolution GED can be generated. A comprehensive comparison is needed to investigate the consistency between the ASTER and ECOSTRESS GEDs before subsequent applications and the release of the ECOSTRESS GED. We conducted a first-ever comparison between the ASTER and ECOSTRESS GEDs in two ways: 1) global spatial comparison and 2) validation using emissivity measurements at 9 sand dune and 2 vegetated sites. The results reveal that the ECOSTRESS emissivities in bands 2 and 4 are in good agreement with ASTER GED, with mean absolute biases (MAB) of ∼ 1 % averaged over the globe. The emissivity difference is the largest for snow and ice in these two bands. For band 5, the discrepancy is relatively larger with a MAB of 1.7 %. The largest emissivity difference is found over evergreen broadleaf forests in humid regions. Comparison with measurements reveals an average root-mean-square error (RMSE) of ECOSTRESS emissivity of ∼ 3 % among all the nine sand sites for all the 3 bands. At the two vegetated sites, the average RMSE is ∼ 1.5 %. The absolute biases are below 3 % in all the bands. The average RMSE between ASTER and ECOSTRESS GEDs among all sand dune sites is 1.29 %. The evaluation results demonstrate a good agreement between these two emissivity retrievals overall and high accuracy of the ECOSTRESS GED.

An enhanced integrated water vapour dataset from more than 10 000 global ground-based GPS stations in 2020

Abstract. We developed a high-quality global integrated water vapour (IWV) dataset from 12 552 ground-based global positioning system (GPS) stations in 2020. It consists of 5 min GPS IWV estimates with a total number of 1 093 591 492 data points. The completeness rates of the IWV estimates are higher than 95 % at 7253 (58 %) stations. The dataset is an enhanced version of the existing operational GPS IWV dataset provided by the Nevada Geodetic Laboratory (NGL). The enhancement is reached by employing accurate meteorological information from the fifth generation of European ReAnalysis (ERA5) for the GPS IWV retrieval with a significantly higher spatiotemporal resolution. A dedicated data screening algorithm is also implemented. The GPS IWV dataset has a good agreement with in situ radiosonde observations at 182 collocated stations worldwide. The IWV biases are within ±3.0 kg m−2 with a mean absolute bias (MAB) value of 0.69 kg m−2. The standard deviations (SD) of IWV differences are no larger than 3.4 kg m−2. In addition, the enhanced IWV product shows substantial improvements compared to NGL's operational version, and it is thus recommended for high-accuracy applications, such as research of extreme weather events and diurnal variations of IWV and intercomparisons with other IWV retrieval techniques. Taking the radiosonde-derived IWV as reference, the MAB and SD of IWV differences are reduced by 19.5 % and 6.2 % on average, respectively. The number of unrealistic negative GPS IWV estimates is also substantially reduced by 92.4 % owing to the accurate zenith hydrostatic delay (ZHD) derived by ERA5. The dataset is available at https://doi.org/10.5281/zenodo.6973528 (Yuan et al., 2022).

Simulations of Mesoscale Flow Systems around Dugway Proving Ground Using the WRF Modeling System

It is widely recognized that regions with complex heterogeneous topography and land-use properties produce a variety of diurnal mesoscale and microscale flows, which can be modified or even masked by significant large-scale synoptic forcing. These flows can be produced through both dynamic and thermal-forcing processes. Recent field programs such as the Terrain-induced Rotor Experiment (T-REX), Mountain Terrain Atmospheric Modeling and Observations Program (MATERHORN), and Perdigao have been used to observe and model flow behaviors under different topographical and large-scale meteorological conditions. Using the Advanced research version of the Weather Research and Forecast (WRF-ARW) model, we applied multi-nesting using an interactive one-way nesting approach to resolve to a sub-kilometer inner-grid spacing (0.452 km). Our interest was in the intensive observation period 6 (IOP6) of the Fall 2012 MATERHORN campaign conducted over Dugway Proving Ground (DPG) in Utah. An initial review of the IOP6 suggested that a range of diurnal flows were present, and that a relatively small subset of model setup configurations would be able to capture the general flows of this period. The review also led us to believe that this same subset would be able to capture differences due to variations in choice of model boundary-layer physics, land surface physics, land use/soil type specifications, and larger-scale meteorological conditions. A high model vertical resolution was used, with 90 vertical sigma levels applied. The IOP6 spanned the period of 2012 0800 UTC 14 October–0800 UTC 15 October. Based upon a lack of deep convection and moist microphysics throughout IOP6, we included comparison of planetary boundary layer (PBL) turbulence parameterization schemes even at the sub-kilometer grid spacing. We focused upon the gross model performance over our inner nest; therefore, a detailed comparison of the effects of model horizontal resolution are excluded. For surface parameters of wind and temperature, we compare mean absolute error and bias scores throughout the period at a number of surface meteorological observing sites. We found that despite attention given to the boundary layer turbulence physics, radiation physics and model vertical resolution, the results seemed to indicate more impact from the choices of thermal soil conductivity parameterization, land surface/soil texture category classification (and associated static property-parameter values), and large-scale forcing model. This finding lends support to what other researchers have found related to how these same forcings can exert a strong influence upon mesoscale flows around DPG. Our findings suggest that the two nights of IOP6 offer a pair of excellent consecutive nights to explore many of the forcing features important to local complex terrain flow. The flows of interest in this case included valley, anabatic/katabatic, and playa breeze systems. Subjective evidence was also found to support an influence provided by the modest synoptic northwesterly flow present within the lower troposphere (mainly on the night of 14 October). Follow-on research using the WRF-ARW capability to nest directly from mesoscale-to-LES can leverage IOP6 further. For example, to uncover more detailed and focused aspects of the dynamic and thermodynamic forcings contributing to the DPG diurnal flows.

Identification of small-scale hydropower potential sites using geographic information system and hydrologic modeling technique: Awata river, Genale Dawa basin, Ethiopia

Small-scale hydropower is a significant source of renewable and inexpensive energy. This study was carried out to identify suitable potential sites for small-scale hydropower for the generation of electricity along the Awata River in Ethiopia. Primary and secondary data collected from various sources were used in the present study to achieve its objective. After the consistency of rainfall was checked and areal rainfall was estimated using the Theissen polygon method, the rainfall–runoff​ relationship was simulated using HEC-HMS. The head of each identified site was estimated by using GIS to overlay the DEM and stream network starting at the outlet. To estimate the discharge, the flow duration curve was developed. HEC-HMS performance was evaluated using Nash–Sutcliffe efficiency (NSE), Mean Absolute Error (MAE), Percent Bias (PBIAS), and coefficient of determination (R2). During calibration, the values were 0.82, 4.39, −2.93, and 0.83, and during validation, the values were 0.76, 5.35, 6.25, and 0.76. The identified sites were ranked using an analytical hierarchy process based on a ratio of five criteria. There are twenty-eight sites identified, with estimated power output and energy output of 45.26 MW and 198.79 GWh at 50% and 6.74 MW and 53.23 GWh at 90%, respectively. The overall rank indicated that the selected site code 29 is ranked first and can be used for the implementation of a small hydropower plant. As a result, the design of all the components of small hydropower plants will be required to implement and generate energy for rural areas in future development.

A Coupled Deep Learning Model for Estimating Surface NO2 Levels From Remote Sensing Data: 15‐Year Study Over the Contiguous United States

AbstractThis study proposes a novel two‐step deep learning (DL) model for estimating surface NO2 concentrations using satellite data over the contiguous United States (CONUS) from 2005 to 2019. The first phase of the model uses partial convolutional neural network (PCNN), an advanced DL model that accurately imputes gaps between surface NO2 stations and creates 5,478 daily‐mean NO2 grids (PCNN‐NO2) of the 2005–2019 period over the study area. We then feed the PCNN‐NO2, along with other predictor variables, into a deep neural network (DNN) to estimate surface NO2 levels, achieving exceptional performance with a correlation coefficient of 0.975–0.978, a mean absolute bias of 0.99–1.38 ppb, and a root mean square error of 1.47–1.97 ppb. Spatial cross‐validation results also indicate strong spatial performance of PCNN‐DNN surface NO2 estimates. In addition to its accurate estimates, the PCNN‐DNN model consistently generates estimated NO2 grids without any missing values, improving the quality of various applications such as emission reduction strategies and public health studies. Between 2005 and 2019, the 5,478 daily estimated NO2 grids over the CONUS reveal significant reductions in NO2 levels in 14 major urban environments: Washington D.C. (−43%), New York (−45%), Los Angeles (−38%), Chicago (−25%), Boston (−43%), Houston (−34%), Dallas (−40%), Philadelphia (−41%), Phoenix (−38%), Detroit (−20%), Denver (−23%), Atlanta (−0.7%), Cincinnati (−38%), and Pittsburgh (−56%). Furthermore, the study shows that the denser urban regions that in‐situ stations are installed in, the higher the difference between in‐situ observations and regional‐mean NO2 levels.

The applicability of sunshine-based global solar radiation models modified with meteorological factors for different climate zones of China

With the development of renewable energy, the exploitation and utilization of solar energy resources also need continuous progress, but solar radiation data shortage has become a serious concern. A method for estimating global solar radiation has been developed to address this issue. The sunshine-based model is currently the most widely used model due to its high calculation accuracy and few input parameters. This paper will first review 13 subcategories (8 categories in total) of the global solar radiation prediction model based on sunshine. Subsequently, the astronomical factors were introduced to modify empirical coefficients, and 8 new categories of models based on sunshine rate were introduced. The radiation data from 83 meteorological stations in China was used to train and validate the model, and the performance of the model was evaluated by using evaluation indicators, such as coefficient of determination (R2), root mean square error (RMSE), mean absolute bias error (MABE), mean bias error (MBE), and global performance index (GPI). The results show that the R2 value of the unmodified empirical model is in a range of 0.82–0.99, and the RMSE value is in a range of 0.018–3.09. In contrast, with the introduction of the astronomical factor, the model accuracy improves significantly, and the modified power function model (N3) gains its best performance. The R2 of model N3 is in a range of 0.86–0.99, and the RMSE value is in a range of 0.018–2.62. The R2 increases by 0.49%, while the RMSE value 6.44%. Above all, it does not require the input of other meteorological parameters for predicting the value of global solar radiation.

Influence of sampling frequency and estimation method on phosphorus load uncertainty in the Western Lake Erie Basin, Ohio, USA

Accurate estimates of nutrient loads are necessary to identify critical source areas and quantify the impact of management practices on pollutant export. Previous studies have investigated nutrient load estimate uncertainty, but they often focus on nutrient loads estimated using an interpolation method for large-scale watersheds with short-term datasets. The study objective was to quantify uncertainty in soluble reactive phosphorus (SRP), total phosphorus (TP), and suspended solids (SS) load estimates from two small (<103 km2) agricultural watersheds in the western Lake Erie Basin resulting from different sampling frequencies. Each watershed had high temporal resolution datasets of discharge (15 min) and nutrient concentration (1 to 3 samples per day) collected over a 30-year period (1990–2020). Firstly, SRP, TP, and SS loads were calculated using the high temporal resolution datasets, which was assumed as “true loads”. Secondly, the high temporal concentration data were decomposed to semiweekly, weekly, biweekly, and monthly sampling and annual loads were estimated using four common load estimation methods to assess the effect of sampling frequency and load estimation method on load estimate error. Across the four different methods, the composite method had the lowest relative root mean square and absolute bias, but the rectangular interpolation method was the most precise. However, even with semiweekly sampling, the composite method resulted in an unacceptable level of precision (average imprecision = 39 %), while the interpolation method resulted in an unacceptable bias (average absolute bias = 16 %). Because neither method could provide acceptable accuracy and precision at the lowest decrease in sampling (e.t. semiweekly sampling), continued daily sampling is recommended in these watersheds.

High-quality PET image synthesis from ultra-low-dose PET/MRI using bi-task deep learning.

Lowering the dose for positron emission tomography (PET) imaging reduces patients' radiation burden but decreases the image quality by increasing noise and reducing imaging detail and quantifications. This paper introduces a method for acquiring high-quality PET images from an ultra-low-dose state to achieve both high-quality images and a low radiation burden. We developed a two-task-based end-to-end generative adversarial network, named bi-c-GAN, that incorporated the advantages of PET and magnetic resonance imaging (MRI) modalities to synthesize high-quality PET images from an ultra-low-dose input. Moreover, a combined loss, including the mean absolute error, structural loss, and bias loss, was created to improve the trained model's performance. Real integrated PET/MRI data from 67 patients' axial heads (each with 161 slices) were used for training and validation purposes. Synthesized images were quantified by the peak signal-to-noise ratio (PSNR), normalized mean square error (NMSE), structural similarity (SSIM), and contrast noise ratio (CNR). The improvement ratios of these four selected quantitative metrics were used to compare the images produced by bi-c-GAN with other methods. In the four-fold cross-validation, the proposed bi-c-GAN outperformed the other three selected methods (U-net, c-GAN, and multiple input c-GAN). With the bi-c-GAN, in a 5% low-dose PET, the image quality was higher than that of the other three methods by at least 6.7% in the PSNR, 0.6% in the SSIM, 1.3% in the NMSE, and 8% in the CNR. In the hold-out validation, bi-c-GAN improved the image quality compared to U-net and c-GAN in both 2.5% and 10% low-dose PET. For example, the PSNR using bi-C-GAN was at least 4.46% in the 2.5% low-dose PET and at most 14.88% in the 10% low-dose PET. Visual examples also showed a higher quality of images generated from the proposed method, demonstrating the denoising and improving ability of bi-c-GAN. By taking advantage of integrated PET/MR images and multitask deep learning (MDL), the proposed bi-c-GAN can efficiently improve the image quality of ultra-low-dose PET and reduce radiation exposure.

Topographic controls on active microwave behavior of mountain snowpacks

An uncalibrated distributed multiphysics snow model driven by downscaled weather forecasts (30-m, 15-min) was implemented as a Radar Observing System Simulator (ROSS) in Senator Beck Basin (SBB), Colorado to elucidate topographic controls on C-, X- and Ku-bands active microwave sensing of mountain snowpacks. Phase-space maps of time-evolving grid-scale ROSS volume backscatter show the accumulation branch of the backscatter-snow water equivalent (σ-SWE) hysteresis seasonal loop that is the physical basis for radar retrieval (direct inference) of SWE and snowpack physical properties. There is good agreement in the accumulation season (R2 ∼ 0.7) between Sentinel-1 and ROSS predictions corrected using average Sentinel-1 measurements under snow free conditions to estimate snow-ground backscatter, capturing well spatial patterns tied to elevation, slope, and aspect. Root Mean Square Deviations (RMSDs) do not exceed ±3.2 dB for ripening snowpacks in early spring and ±2.4 dB for dry snowpacks in the accumulation season when the mean absolute bias is <1 dB for all land-cover types with topographic slopes ≤ 30°. Grid-point RMSDs are attributed to the underestimation of snowfall on upwind slopes compounded with forecast errors for the weather near the ground. Like Sentinel-1, ROSS backscatter fields exhibit frequency-independent single-scaling behavior in the 60–150 m scale range for dry snowpacks in the accumulation season, while frequency-dependent scaling behavior emerges in the ablation season. This study demonstrates skillful physical modeling capabilities to emulate Sentinel-1 observations in complex terrain. Conversely, it suggests high readiness to retrieve snow mass and snowpack properties in mountainous regions from radar measurements at high-spatial resolutions enabled by SAR technology.

Inferring the linkage of sea surface height anomalies, surface wind stress and sea surface temperature with the falling ice radiative effects using satellite data and global climate models

This study attempts to infer the linkage of sea surface height anomaly (SSHA), surface wind stress and sea surface temperature with the falling ice (snow) radiative effects (FIREs) over the tropical and subtropical Pacific Ocean using CESM1-CAM5 sensitivity experiments with FIREs-off (NOS) and on (SON) under CMIP5 historical run. The obs4MIPs monthly SSH data based upon satellite measurements are used as a reference. The seasonal and annual mean spatial patterns of SSHA difference between NOS and SON are tightly linked to those of SST and TAU over the study domain, in particular, over the south Pacific. Compared with NOS, SON simulates improved seasonal and annual mean SSHA associated with improved sea surface temperature (SST), surface wind stress (TAU) over the trade-wind areas. In SON, the simulated mean absolute bias of SSHA over the study domain is reduced (up to 30%) against NOS relative to observations. The SSHA biases are then compared with CMIP5 models. Despite the biases of SST and SSHA over the south and north flanks of the equator in SON, the seasonal variations of improved SSHA are closely related to those of TAU and SST resulting from the FIREs; that is, higher SSHA is associated with weaker TAU and warmer SST changes and vice versa. The CMIP5 ensemble mean absolute biases of SSHA show similarities to NOS mainly over the south Pacific.

New parameterization of air-sea exchange coefficients and its impact on intensity prediction under major tropical cyclones

Understanding and quantifying air-sea exchanges of enthalpy and momentum fluxes are crucial for the advanced prediction of tropical cyclone (TC) intensity. Here, we present a new parameterization of air-sea fluxes at extreme wind speeds from 40 m s−1 to 75 m s−1, which covers the range of major TCs. Our approach assumes that the TC can reach its maximum potential intensity (MPI) if there are no influences of external forces such as vertical wind shear or other environmental constraints.This method can estimate the ratio of the enthalpy and momentum exchange coefficient (Ck/Cd) under the most intense TCs without direct flux measurements. The estimation showed that Ck/Cd increases with wind speed at extreme winds above 40 m s−1. Two types of surface layer schemes of the Hurricane Weather and Research Forecast (HWRF) were designed based on the wind speed dependency of the Ck/Cd found at high winds: (i) an increase of Ck/Cd based on decreasing Cd (Cd_DC) and (ii) an increase of Ck/Cd based on increasing Ck (Ck_IC). The modified surface layer schemes were compared to the original HWRF scheme (using nearly fixed Cd and Ck at extreme winds; CTRL) through idealized experiments and real-case predictions. The idealized experiments showed that Cd_DC reduced frictional dissipation in the air-sea interface as well as significantly reduced sea surface cooling, making the TC stronger than other schemes. As a result, Cd_DC reduced the mean absolute error and negative bias by 15.0% (21.0%) and 19.1% (32.0%), respectively, for all lead times of Hurricane Irma in 2017 (Typhoon Mangkhut in 2018) compared to CTRL. This result suggests that new parameterization of Ck/Cd with decreasing Cd at high winds can help improve TC intensity prediction, which currently suffers from underestimating the intensity of the strongest TCs.