Thin-plate Smoothing Spline Research Articles

Predicting Mortality Risks Using Body Mass Index and Weight Loss at Admission in Patients with Heart Failure.

The association of the combination of body mass index (BMI) and weight change at admission with prognoses in patients with heart failure (HF) is unclear. Therefore, we investigated whether BMI and weight changes at admission affect mortality in patients with HF. This retrospective cohort study lasted 99 months, starting in April 2014, and included 4,862 patients with HF from a Japanese real-world database. Cubic and thin-plate smoothing spline analyses were performed to investigate the association of BMI and weight changes with mortality. The percentage weight change was calculated every 6 months. The study outcome was the presence or absence of death. The patients' mean age was 81.5±9.6 years, and 1,239 (25.5%) patients died. Cubic spline analysis revealed a negative correlation of BMI with mortality hazard ratio (HR) (BMI of 18.5 kg/m2 and 25 kg/m2; HR=1.3 [1.2-1.4] and 0.8 [0.7-0.9], respectively). Cubic spline analysis of weight change showed that weight loss tended to increase the mortality HR (each 6% decrease in weight change rate was associated with a 1.1 times higher mortality risk (95% CI [1.0-1.2]) Thin-plate smoothing spline analysis showed that the odds ratio (OR) negatively correlated with BMI (1-year mortality: BMI of 18.5 kg/m2, 22 kg/m2, and 25 kg/m2; OR at 0% weight change=1.5, 1.0, and 0.7, respectively; 2-year mortality: BMI=18.5 kg/m2, 22 kg/m2, and 25 kg/m2; OR at 0% weight change=1.4, 0.9, and 0.7, respectively). A low BMI in patients with HF was associated with a higher risk of mortality. Weight loss in patients, regardless of BMI, was associated with a higher OR for mortality.

Development of residual stress model of orthotropic steel decks using measurements

Orthotropic steel decks are widely used in the bridge engineering practice due to light weight, high performance and high ultimate strength-to-weight ratio. Purpose of the current paper is developing the first residual stress model for orthotropic steel decks (OSDs) with U-shaped stiffeners based on residual stress measurement results. The proposed simplified residual stress model is valid for OSDs with single-bevel butt welds, manufactured from steel grade up to S355. Distribution and magnitude of manufacturing-induced residual stresses due to hot rolling, thermal cutting, cold-forming, welding, etc. are determined using contour method with thin-plate smoothing spline. It makes it possible to interpret two-dimensional full-map residual stresses, instead of conventional pointwise residual stress measurement techniques, resulting in higher spatial resolution. Longitudinal residual stress patterns are evaluated in representative cross-sections clearly indicating high stress peaks near the welds, even exceeding yield strength of the base material, while the influence of thermal cutting and cold-forming can be addressed as well. Statistical evaluation of measurement results is carried out, which resulted in mean standard deviation of 22 MPa and 34 MPa in the stiffener and deck plate, respectively, which is comparable with the accuracy of other measurement techniques, e.g., X-ray diffraction method or hole-drilling technique.

Spatiotemporal variability characteristics of extreme climate events in Xinjiang during 1960–2019

Under the global warming, it is particularly important to explore the response of extreme climate to global climate change over the arid regions. Based on daily temperature (maximum, minimum, and average) and precipitation data from meteorological stations in Xinjiang, China, we analyzed the spatiotemporal characteristics of extreme temperature and extreme precipitation events via combining thin plate smoothing spline function interpolation, Sen's slope, and Mann-Kendall test. Our results showed that during 1960-2019, the extreme low temperature index of frost days (FD), icing days (ID), cold days (TX10p), cold nights (TN10p), and cold speel duration index (CSDI) all showed the downward trend to varying degrees, and the extreme high temperature index of summer days (SD25), warm days (TX90p), warm night (TN90p), and warm speel duration index (WSDI) all showed an upward trend to varying degrees, and the extreme low temperature index of high altitude mountains decreases more than that of the basin and plains. In addition, all the extreme temperature indices are closely related to the annual average temperature in Xinjiang (R > 0.6). Among the extreme precipitation indices, except for the consecutive dry days (CDD), the other extreme precipitation indices showed increasing trends to different degrees, but the changes in extreme precipitation in Xinjiang were mainly manifested by the increase of heavy precipitation in a short period (the increase of heavy precipitation and extreme heavy precipitation was the largest, 44.8mm/10a and 17.6mm/10a, respectively) and spatially concentrated in the Ili River and Altai Mountains in northern Xinjiang. Meanwhile, annual precipitation was positively correlated with the extreme precipitation index (R > 0.4), except for the CDD. This study provides theoretical support for the prevention and control of natural disasters in the dry zone.

Daily High-Resolution Temperature and Precipitation Fields for the Contiguous United States from 1951 to Present

Abstract In this paper, a new set of daily gridded fields and area averages of temperature and precipitation is introduced that covers the contiguous United States (CONUS) from 1951 to present. With daily updates and a grid resolution of approximately 0.0417° (nominally 5 km), the product, named nClimGrid-Daily, is designed to be used particularly in climate monitoring and other applications that rely on placing event-specific meteorological patterns into a long-term historical context. The gridded fields were generated by interpolating morning and midnight observations from the Global Historical Climatology Network–Daily dataset using thin-plate smoothing splines. Additional processing steps limit the adverse effects of spatial and temporal variations in station density, observation time, and other factors on the quality and homogeneity of the fields. The resulting gridded data provide smoothed representations of the point observations, although the accuracy of estimates for individual grid points and days can be sensitive to local spatial variability and the ability of the available observations and interpolation technique to capture that variability. The nClimGrid-Daily dataset is therefore recommended for applications that require the aggregation of estimates in space and/or time, such as climate monitoring analyses at regional to national scales. Significance Statement Many applications that use historical weather observations require data on a high-resolution grid that are updated daily. Here, a new dataset of daily temperature and precipitation for 1951–present is introduced that was created by interpolating irregularly spaced observations to a regular grid with a spacing of 0.0417° across the contiguous United States. Compared to other such datasets, this product is particularly suitable for monitoring climate and drought on a daily basis because it was processed so as to limit artificial variations in space and time that may result from changes in the types and distribution of observations used.

Seasonal Changes in Climate Variables in Rainfed Crop Areas in the Lerma-Chapala-Santiago Basin, Mexico

This paper shows the effects of changes in the spatial-temporal behavior and phase shift of climate variables on rainfed agriculture in the Lerma-Chapala-Santiago Basin in central Mexico. Specifically, changes in rainfall (R), maximum temperature (Tmax), and minimum temperature (Tmin) were analyzed over two 25-year periods (1960 to 1985 and 1986 to 2010). Climate surfaces were generated by interpolation using the thin-plate smoothing spline algorithm in the software ANUSPLIN. Climate data were Fourier-transformed and fitted to a sinusoidal curve model, and changes in amplitude (increase) and phase were analyzed. The temporal behavior (1960–2010) indicated that rainfall was the most stable variable at the monthly level and presented no significant changes. However, Tmax increased by 2°C in the final period, and Tmin increased by 0.7°C at the end of the final period. The basin was discretized into ten rainfed crop areas (RCAs) according to the extent of changes in the amplitude and phase of the climate variables. The central and southern portions (55% of the area) presented more significant changes in amplitude, mainly in Tmin and Tmax. The remaining RCAs were smaller (14.6%) but presented greater variation: the amplitude of the Tmin decreased in addition to showing a phase shift, whereas Tmax increased in addition to showing a phase shift. These results translate into a delay in the characteristic temperatures of the spring and summer seasons, which can impact the rainfed crop cycle. Additionally, rainfall showed an annual decrease of approximately 50 mm in all RCAs, which can affect the phenological development of crops during critical stages (emergence through flowering). These changes represent a significant threat to the regional economy and food security of Mexico.

Assessment of Snowmaking Conditions Based on Meteorological Reconstruction in the Beijing–Zhangjiakou Mountain Area of North China in 1978–2017

AbstractThe outdoor events of the 2022 Winter Olympics and Paralympics will be held in the mountain areas of Beijing–Zhangjiakou, North China, where there is a complete reliance on artificial snow production owing to the dry and cold weather conditions. To assess how favorable the meteorological conditions are to snowmaking at the mountain venues, we reconstructed the daily wet-bulb temperature by adopting the thin-plate smoothing spline function method, and then we assessed the potential number of snowmaking days at eight weather stations (928–2098 m MSL) from October to the next April (i.e., the ski season) during the period 1978–2017. Results showed that artificial snow production would have been possible on 121 (±14) to 171 (±12) days on average at the stations with the increases of altitude, and the number of days decreased at rates of 4.3–5.1 days decade−1 across four decades of the study period. The cause of the decrease was the warming trend, which affected the number of days in low-altitude sites simultaneously, but the reduction was delayed with increased elevation. At monthly scale, the number of snowmaking days was robust in wintertime but reduced in other months of the ski season, particularly in March in more recent subperiods at high-altitude stations, which was determined by the increase in high values of daily mean wet-bulb temperature. Further improvements in assessing snowmaking conditions require detailed microclimatic studies to reduce the uncertainties caused by meteorological conditions as well as combination with model-based methods to determine potential future changes.

An ecological analysis of long-term exposure to PM2.5 and incidence of COVID-19 in Canadian health regions

BackgroundAmbient fine particulate matter (PM2.5) is associated with a wide range of acute and chronic health effects, including increased risk of respiratory infection. However, evidence specifically related to novel coronavirus disease (COVID-19) is limited. MethodsCOVID-19 case counts for 111 Canadian health regions were obtained from the COVID-19 Canada Open Data portal. Annual PM2.5 data for 2000–2016 were estimated from a national exposure surface based on remote sensing, chemical transport modelling and ground observations, and minimum and maximum temperature data for 2000–2015 were based on a national interpolated surface derived from thin-plate smoothing splines. Population counts and sociodemographic data by health region were obtained from the 2016 census, and health data (self-rated health and prevalence of smoking, obesity, and selected chronic diseases) by health region, were obtained from the Canadian Community Health Survey. Data on total number of COVID-19 tests and changes in mobility comparing post-vs. pre-introduction of social distancing measures were available by province. Data were analyzed using negative binomial regression models. ResultsAfter controlling for province, temperature, demographic and health characteristics and days since peak incidence by health region, long-term PM2.5 exposure exhibited a positive association with COVID-19 incidence (incidence rate ratio 1.07, 95% confidence interval 0.97–1.18 per μg/m3). This association was larger in magnitude and statistically significant in analyses excluding provinces that reported cases only for aggregated health regions, excluding health regions with less than median population density, and restricted to the most highly affected provinces (Quebec and Ontario). ConclusionsWe observed a positive association between COVID-19 incidence and long-term PM2.5 exposure in Canadian health regions. The association was larger in magnitude and statistically significant in more highly affected health regions and those with potentially less exposure measurement error. While our results generate hypotheses for further testing, they should be interpreted with caution and require further examination using study designs less prone to bias.

Spatio‐temporal temperature ­variations in the Chinese Yellow River basin from 1981 to 2013

Air temperature changes are one of the main factors that cause variability in the natural landscape and living environment. In this study, spatio‐temporal evolutions in temperature in the Yellow River basin were statistically interpolated by the partial thin plate smoothing spline method for the period 1981–2013. Results indicate that: (1) the interpolation method can be successfully applied to simulate the spatio‐temporal evolution of meteorological data in large regions with complex terrain; (2) annual temperature distributions show no obvious changes from 1981 to 2013 (the spatial distribution maintains the form of an overall decreasing trend from the east to the west of the basin ‐ the annual average temperature in the east of the basin is about 13–15°C, while in the west it is about –7 to –3°C); (3) the average lowest and highest temperatures in January and December were concentrated in the western and southeastern regions of the basin respectively; January has the lowest temperature, while July has the highest temperature; the lowest temperature from May to September may stay above 0°C, and in the remaining months it will stay below 0°C. The research method applied in this study can provide a reference for the interpolation of meteorological data in other study areas.

Evaluating uncertainties in modelling the snow hydrology of the Fraser River Basin, British Columbia, Canada

Abstract. This study evaluates predictive uncertainties in the snow hydrology of the Fraser River Basin (FRB) of British Columbia (BC), Canada, using the Variable Infiltration Capacity (VIC) model forced with several high-resolution gridded climate datasets. These datasets include the Canadian Precipitation Analysis and the thin-plate smoothing splines (ANUSPLIN), North American Regional Reanalysis (NARR), University of Washington (UW) and Pacific Climate Impacts Consortium (PCIC) gridded products. Uncertainties are evaluated at different stages of the VIC implementation, starting with the driving datasets, optimization of model parameters, and model calibration during cool and warm phases of the Pacific Decadal Oscillation (PDO). The inter-comparison of the forcing datasets (precipitation and air temperature) and their VIC simulations (snow water equivalent – SWE – and runoff) reveals widespread differences over the FRB, especially in mountainous regions. The ANUSPLIN precipitation shows a considerable dry bias in the Rocky Mountains, whereas the NARR winter air temperature is 2 °C warmer than the other datasets over most of the FRB. In the VIC simulations, the elevation-dependent changes in the maximum SWE (maxSWE) are more prominent at higher elevations of the Rocky Mountains, where the PCIC-VIC simulation accumulates too much SWE and ANUSPLIN-VIC yields an underestimation. Additionally, at each elevation range, the day of maxSWE varies from 10 to 20 days between the VIC simulations. The snow melting season begins early in the NARR-VIC simulation, whereas the PCIC-VIC simulation delays the melting, indicating seasonal uncertainty in SWE simulations. When compared with the observed runoff for the Fraser River main stem at Hope, BC, the ANUSPLIN-VIC simulation shows considerable underestimation of runoff throughout the water year owing to reduced precipitation in the ANUSPLIN forcing dataset. The NARR-VIC simulation yields more winter and spring runoff and earlier decline of flows in summer due to a nearly 15-day earlier onset of the FRB springtime snowmelt. Analysis of the parametric uncertainty in the VIC calibration process shows that the choice of the initial parameter range plays a crucial role in defining the model hydrological response for the FRB. Furthermore, the VIC calibration process is biased toward cool and warm phases of the PDO and the choice of proper calibration and validation time periods is important for the experimental setup. Overall the VIC hydrological response is prominently influenced by the uncertainties involved in the forcing datasets rather than those in its parameter optimization and experimental setups.

Correcting bias in the rational polynomial coefficients of satellite imagery using thin-plate smoothing splines

The Rational Function Model (RFM) has proven to be a viable alternative to the rigorous sensor models used for geo-processing of high-resolution satellite imagery. Because of various errors in the satellite ephemeris and instrument calibration, the Rational Polynomial Coefficients (RPCs) supplied by image vendors are often not sufficiently accurate, and there is therefore a clear need to correct the systematic biases in order to meet the requirements of high-precision topographic mapping. In this paper, we propose a new RPC bias-correction method using the thin-plate spline modeling technique. Benefiting from its excellent performance and high flexibility in data fitting, the thin-plate spline model has the potential to remove complex distortions in vendor-provided RPCs, such as the errors caused by short-period orbital perturbations. The performance of the new method was evaluated by using Ziyuan-3 satellite images and was compared against the recently developed least-squares collocation approach, as well as the classical affine-transformation and quadratic-polynomial based methods. The results show that the accuracies of the thin-plate spline and the least-squares collocation approaches were better than the other two methods, which indicates that strong non-rigid deformations exist in the test data because they cannot be adequately modeled by simple polynomial-based methods. The performance of the thin-plate spline method was close to that of the least-squares collocation approach when only a few Ground Control Points (GCPs) were used, and it improved more rapidly with an increase in the number of redundant observations. In the test scenario using 21 GCPs (some of them located at the four corners of the scene), the correction residuals of the thin-plate spline method were about 36%, 37%, and 19% smaller than those of the affine transformation method, the quadratic polynomial method, and the least-squares collocation algorithm, respectively, which demonstrates that the new method can be more effective at removing systematic biases in vendor-supplied RPCs.

Spatial interpolation techniques: their applications in regionalizing climate-change series and associated accuracy evaluation in Northeast China

ABSTRACTAn accurate gridded climatological temperature data-set can be a reliable basis for studying the issues that concern climate change, natural disasters, among others. In this study, the climate standard value data and annual observed data collected by 104 meteorological sites in Northeast China from 1971 to 2000 are used to interpolate the annual mean air temperature of the region with different spatial interpolation techniques. Efforts have also been made to verify the accuracy of the results generated by each interpolation method with error indicator, temperature characteristic value and temperature variation curves. The results show at the spatial scale, the partial thin-plate smoothing spline (PTPSS) scheme generates the most desirable temperature interpolations with a root-mean-square error at 0.34 °C and an average standard error at 0.52 °C. At the temporal scale, the PTPSS-based 1971–2000 temperature curves agree the best with the gridded temperature curves, with the correlation coefficients of means, minimums and maximums between the two being larger than 0.9. Meanwhile, the PTPSS produce the smallest error approaching zero among all the interpolation schemes tested in the study. In this context, the PTPSS scheme stands out as the most desirable interpolation method for the annual mean temperature in Northeast China.

Comparing interpolation techniques for annual temperature mapping across Xinjiang region

Interpolating climatic variables such as temperature is challenging due to the highly variable nature of meteorological processes and the difficulty in establishing a representative network of stations. In this paper, based on the monthly temperature data which obtained from the 154 official meteorological stations in the Xinjiang region and surrounding areas, we compared five spatial interpolation techniques: Inverse distance weighting (IDW), Ordinary kriging, Cokriging, thin-plate smoothing splines (ANUSPLIN) and Empirical Bayesian kriging(EBK). Error metrics were used to validate interpolations against independent data. Results indicated that, the ANUSPLIN performed best than the other four interpolation methods.

Can monthly precipitation interpolation error be reduced by adding periphery climate stations? A case study in China's land border areas

Minor interpolation error of spatially continuous precipitation is increasingly in demand to support many climate studies. In this paper, based on the thin-plate smoothing splines (ANUSPLIN), we studied the effects of adding periphery stations on monthly precipitation interpolation errors in China with 184 stations from neighboring countries during 1971–2000. Here, we show that with the exception of the northern piedmont of the Himalayas, the interpolation accuracy of monthly precipitation was improved greatly in China's border areas. Mean absolute error was reduced by an average of 2.8 mm month−1 across 21 withheld stations. By incorporating 184 foreign stations into interpolation, the overestimated precipitation in the northern piedmont of the Himalayas can be primarily attributed to the drawback that ANUSPLIN had difficulty estimating sharply varying rain shadows in the Qinghai–Tibetan Plateau. Overall, these results mentioned above emphasized the importance of periphery stations to generate gridded precipitation datasets and the limitation of ANUSPLIN to simulate terrain-induced climate transitions.

Countervailing effects of income, air pollution, smoking, and obesity on aging and life expectancy: population-based study of U.S. Counties.

BackgroundIncome, air pollution, obesity, and smoking are primary factors associated with human health and longevity in population-based studies. These four factors may have countervailing impacts on longevity. This analysis investigates longevity trade-offs between air pollution and income, and explores how relative effects of income and air pollution on human longevity are potentially influenced by accounting for smoking and obesity.MethodsCounty-level data from 2,996 U.S. counties were analyzed in a cross-sectional analysis to investigate relationships between longevity and the four factors of interest: air pollution (mean 1999–2008 PM2.5), median income, smoking, and obesity. Two longevity measures were used: life expectancy (LE) and an exceptional aging (EA) index. Linear regression, generalized additive regression models, and bivariate thin-plate smoothing splines were used to estimate the benefits of living in counties with higher incomes or lower PM2.5. Models were estimated with and without controls for smoking, obesity, and other factors.ResultsModels which account for smoking and obesity result in substantially smaller estimates of the effects of income and pollution on longevity. Linear regression models without these two variables estimate that a $1,000 increase in median income (1 μg/m3 decrease in PM2.5) corresponds to a 27.39 (33.68) increase in EA and a 0.14 (0.12) increase in LE, whereas models that control for smoking and obesity estimate only a 12.32 (20.22) increase in EA and a 0.07 (0.05) increase in LE. Nonlinear models and thin-plate smoothing splines also illustrate that, at higher levels of income, the relative benefits of the income-pollution tradeoff changed—the benefit of higher incomes diminished relative to the benefit of lower air pollution exposure.ConclusionsHigher incomes and lower levels of air pollution both correspond with increased human longevity. Adjusting for smoking and obesity reduces estimates of the benefits of higher income and lower air pollution exposure. This adjustment also alters the tradeoff between income and pollution: increases in income become less beneficial relative to a fixed reduction in air pollution—especially at higher levels of income.

A 1973–2008 Archive of Climate Surfaces for NW Maghreb

Climate archives are time series. They are used to assess temporal trends of a climate-dependent target variable, and to make climate atlases. A high-resolution gridded dataset with 1728 layers of monthly mean maximum, mean and mean minimum temperatures and precipitation for the NW Maghreb (28°N–37.3°N, 12°W–12°E, ~1-km resolution) from 1973 through 2008 is presented. The surfaces were spatially interpolated by ANUSPLIN, a thin-plate smoothing spline technique approved by the World Meteorological Organization (WMO), from georeferenced climate records drawn from the Global Surface Summary of the Day (GSOD) and the Global Historical Climatology Network-Monthly (GHCN-Monthly version 3) products. Absolute errors for surface temperatures are approximately 0.5 °C for mean and mean minimum temperatures, and peak up to 1.76 °C for mean maximum temperatures in summer months. For precipitation, the mean absolute error ranged from 1.2 to 2.5 mm, but very low summer precipitation caused relative errors of up to 40% in July. The archive successfully captures climate variations associated with large to medium geographic gradients. This includes the main aridity gradient which increases in the S and SE, as well as its breaking points, marked by the Atlas mountain range. It also conveys topographic effects linked to kilometric relief mesoforms.

Local Adaptive Calibration of the Satellite-Derived Surface Incident Shortwave Radiation Product Using Smoothing Spline

Incident solar radiation $(R_s)$ over the Earth's surface plays an important role in determining the Earth's climate and environment. Generally, $R_s$ can be obtained from direct measurements, remotely sensed data, or reanalysis and general circulation model (GCM) data. Each type of product has advantages and limitations: the surface direct measurements provide accurate but sparse spatial coverage, whereas other global products may have large uncertainties. Ground measurements have been normally used for validation and occasionally calibration, but transforming their “true values” spatially to improve the satellite products is still a new and challenging topic. In this paper, an improved thin-plate smoothing spline approach is presented to locally “calibrate” the Global LAnd Surface Satellite (GLASS) $R_s$ product using the reconstructed $R_s$ data from surface meteorological measurements. The influence of surface elevation on $R_s$ estimation was also considered in the proposed method. The point-based surface reconstructed $R_s$ was used as the response variable, and the GLASS $R_s$ product and the surface elevation data at the corresponding locations as explanatory variables to train the thin-plate spline model. We evaluated the performance of the approach using the cross-validation method at both daily and monthly time scales over China. We also validated the estimated $R_s$ based on the thin-plate spline method using independent ground measurements and independent satellite estimates of $R_s$ . These validation results indicated that the thin-plate smoothing spline method can be effectively used for calibrating satellite-derived $R_s$ products using ground measurements to achieve better accuracy.

Spatial Analysis of Renewable Energy in Papua New Guinea through Remote Sensing and GIS

Electrification and sustainable energy uses are increasing in Papua New Guinea (PNG) over the last few decades. The bulk of PNG’s population (85%) lives in isolated and dispersed villages in the rural areas. Most of these isolated and dispersed areas are still yet to be connected to an electricity supply. Papua New Guinea (PNG) is richly endowed with natural resources, but exploitation has been hampered by rugged terrain, land tenure issues, and the high cost of developing infrastructure. The study is focused on mapping of enriched renewable energy zones of the entire country. Different variables related to renewable, like surface albedo index, earth skin temperature, solar insolation incident, and wind speed are used for this purpose. Three interpolation approaches, like inverse distance weighted averaging, thin-plate smoothing splines, and kriging, are evaluated to interpolate all variables. Rating and weight sum overlay operation is applied to derive potential renewable energy zones in this equatorial country. Results show that potential renewable energy distribution is high in Papua New Guinea on the March and September equinoxes. Yearly average distribution of renewable energy source variables is significantly higher in most areas of Manus, New Ireland, North Solomon, West New Britain, Northern, Central and Milne Bay; a larger portion of East New Britain; the northern part of West and East Sepik, Central, Morobe and eastern part of Madang province. The potential renewable energy distribution data can help to establish sustainable energy production in the country.

Carbon carry capacity and carbon sequestration potential in China based on an integrated analysis of mature forest biomass.

Forests play an important role in acting as a carbon sink of terrestrial ecosystem. Although global forests have huge carbon carrying capacity (CCC) and carbon sequestration potential (CSP), there were few quantification reports on Chinese forests. We collected and compiled a forest biomass dataset of China, a total of 5841 sites, based on forest inventory and literature search results. From the dataset we extracted 338 sites with forests aged over 80 years, a threshold for defining mature forest, to establish the mature forest biomass dataset. After analyzing the spatial pattern of the carbon density of Chinese mature forests and its controlling factors, we used carbon density of mature forests as the reference level, and conservatively estimated the CCC of the forests in China by interpolation methods of Regression Kriging, Inverse Distance Weighted and Partial Thin Plate Smoothing Spline. Combining with the sixth National Forest Resources Inventory, we also estimated the forest CSP. The results revealed positive relationships between carbon density of mature forests and temperature, precipitation and stand age, and the horizontal and elevational patterns of carbon density of mature forests can be well predicted by temperature and precipitation. The total CCC and CSP of the existing forests are 19.87 and 13.86 Pg C, respectively. Subtropical forests would have more CCC and CSP than other biomes. Consequently, relying on forests to uptake carbon by decreasing disturbance on forests would be an alternative approach for mitigating greenhouse gas concentration effects besides afforestation and reforestation.

Mapping near-surface air temperature, pressure, relative humidity and wind speed over Mainland China with high spatiotemporal resolution

As part of a joint effort to construct an atmospheric forcing dataset for mainland China with high spatiotemporal resolution, a new approach is proposed to construct gridded near-surface temperature, relative humidity, wind speed and surface pressure with a resolution of 1 km×1 km. The approach comprises two steps: (1) fit a partial thin-plate smoothing spline with orography and reanalysis data as explanatory variables to ground-based observations for estimating a trend surface; (2) apply a simple kriging procedure to the residual for trend surface correction. The proposed approach is applied to observations collected at approximately 700 stations over mainland China. The generated forcing fields are compared with the corresponding components of the National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis dataset and the Princeton meteorological forcing dataset. The comparison shows that, both within the station network and within the resolutions of the two gridded datasets, the interpolation errors of the proposed approach are markedly smaller than the two gridded datasets.

Performance criteria and discrimination of extreme undersmoothing in nonparametric regression

The prediction error (average squared error) is the most commonly used performance criterion for the assessment of nonparametric regression estimators. However, there has been little investigation of the properties of the criterion itself. This paper shows that in certain situations the prediction error can be very misleading because it fails to discriminate an extreme undersmoothed estimate from a good estimate. For spline smoothing, we show, using asymptotic analysis and simulations, that there is poor discrimination of extreme undersmoothing in the following situations: small sample size or small error variance or a function with high curvature. To overcome this problem, we propose using the Sobolev error criterion. For spline smoothing, it is shown asymptotically and by simulations that the Sobolev error is significantly better than the prediction error in discriminating extreme undersmoothing. Similar results hold for other nonparametric regression estimators and for multivariate smoothing. For thin-plate smoothing splines, the prediction error׳s poor discrimination of extreme undersmoothing becomes significantly worse with increasing dimension.