The purpose of the study was to develop pedotransfer functions for determining saturated hydraulic conductivity (Ks). Pedotransfer functions (PTFs) for predicting soil physical properties used in determining saturated...The purpose of the study was to develop pedotransfer functions for determining saturated hydraulic conductivity (Ks). Pedotransfer functions (PTFs) for predicting soil physical properties used in determining saturated hydraulic conductivity, based on moisture retention characteristics, were developed. The van Genuchten moisture retention equation was fitted to measured moisture retention properties obtained from International Soil Reference and Information Centre (ISRIC) soils data base in order to determine parameters in the equation i.e. saturated soil moisture content (θs), residual soil moisture (θr), air entry parameter (α) and the pore size distribution parameter (n). 457 samples drawn from the data base were used to be the maximum possible sample size that contained the measured soils characteristics data required. Using statistical regression, mathematical relationships were developed between moisture retention parameters (response variables) and appropriately selected transformed basic soil properties (predictor variables). The developed PTFs were evaluated for accuracy and reliability. It was found that pedotransfer functions developed for θs produced the best performance in reliability compared to the remaining parameters yielding a correlation coefficient value of coefficient of determination (R2 = 0.76), RMSE = 2.09, NSE = 0.75 and RSR = 0.5 indicating good performance. Relatively poorest performance was obtained from the pedotransfer function developed for α which yielded a correlation coefficient, R2= 0.06, RMSE = 0.85 and a NSE of 0.02 reflecting the best possible equation derived for the parameter for use in predicting hydraulic conductivity. Out of the pedotransfer functions developed for each of the moisture retention parameters, the best performing PTF was identified for each parameter. The accuracy of the pedotransfer functions assessed based on R2 were for θs (R2 = 0.80), θr (R2 = 0.42), α (R2 = 0.04) and for n (R2 = 0.30), when the variables were expressed directly in terms of the s展开更多
We combined C and N related pedon data from the USDA-NRCS National Cooperative Soil Survey Soil Characterization Database with data from the University of Alaska Fairbanks (UAF) northern soils research program, repres...We combined C and N related pedon data from the USDA-NRCS National Cooperative Soil Survey Soil Characterization Database with data from the University of Alaska Fairbanks (UAF) northern soils research program, representing 58 and 30 years of field work, respectively. Carbon and N data from 117 UAF pedons were added to 541 pedons from the USDA-NRCS data set for a total of 658. Missing carbon (C), nitrogen (N) and related data were added to nearly all of the USDA-NRCS Arctic region pedons from unpublished UAF data. We present relationships among soil parameters of the data set that are necessary for calculation of pedon soil organic C and N stores. These new relationships are necessary for better estimating missing soil bulk density (Db) from measured soil organic C by high-temperature combustion (SOCHTC) and for conversion of acid chromate reduction soil organic carbon (SOCACR) to SOCHTC. For the USDA-NRCS data, missing Db data were estimated and SOCACR corrected to SOCHTC using the new functional relationships developed. This allowed for pedon SOC and N stores to be calculated for 609 and 468 Alaska pedons respectively, the most available to-date. Additionally, functional relationships were developed for data within soil orders to estimate total SOCHTC and N stores in pedons with missing surface organic horizons where only thicknesses were known. These relationships are presented in order to fill-in missing data and to better define the existing data set for future use. Some 1904 missing Db data points and 1612 corrected SOCHTC data points were added to the total of 4240 points in the 609 pedons that constitute the updated dataset. When O-layer thickness functions developed here were used, SOC and N stores were calculated for an additional 137 and 184 pedons respectively.展开更多
文摘The purpose of the study was to develop pedotransfer functions for determining saturated hydraulic conductivity (Ks). Pedotransfer functions (PTFs) for predicting soil physical properties used in determining saturated hydraulic conductivity, based on moisture retention characteristics, were developed. The van Genuchten moisture retention equation was fitted to measured moisture retention properties obtained from International Soil Reference and Information Centre (ISRIC) soils data base in order to determine parameters in the equation i.e. saturated soil moisture content (θs), residual soil moisture (θr), air entry parameter (α) and the pore size distribution parameter (n). 457 samples drawn from the data base were used to be the maximum possible sample size that contained the measured soils characteristics data required. Using statistical regression, mathematical relationships were developed between moisture retention parameters (response variables) and appropriately selected transformed basic soil properties (predictor variables). The developed PTFs were evaluated for accuracy and reliability. It was found that pedotransfer functions developed for θs produced the best performance in reliability compared to the remaining parameters yielding a correlation coefficient value of coefficient of determination (R2 = 0.76), RMSE = 2.09, NSE = 0.75 and RSR = 0.5 indicating good performance. Relatively poorest performance was obtained from the pedotransfer function developed for α which yielded a correlation coefficient, R2= 0.06, RMSE = 0.85 and a NSE of 0.02 reflecting the best possible equation derived for the parameter for use in predicting hydraulic conductivity. Out of the pedotransfer functions developed for each of the moisture retention parameters, the best performing PTF was identified for each parameter. The accuracy of the pedotransfer functions assessed based on R2 were for θs (R2 = 0.80), θr (R2 = 0.42), α (R2 = 0.04) and for n (R2 = 0.30), when the variables were expressed directly in terms of the s
文摘We combined C and N related pedon data from the USDA-NRCS National Cooperative Soil Survey Soil Characterization Database with data from the University of Alaska Fairbanks (UAF) northern soils research program, representing 58 and 30 years of field work, respectively. Carbon and N data from 117 UAF pedons were added to 541 pedons from the USDA-NRCS data set for a total of 658. Missing carbon (C), nitrogen (N) and related data were added to nearly all of the USDA-NRCS Arctic region pedons from unpublished UAF data. We present relationships among soil parameters of the data set that are necessary for calculation of pedon soil organic C and N stores. These new relationships are necessary for better estimating missing soil bulk density (Db) from measured soil organic C by high-temperature combustion (SOCHTC) and for conversion of acid chromate reduction soil organic carbon (SOCACR) to SOCHTC. For the USDA-NRCS data, missing Db data were estimated and SOCACR corrected to SOCHTC using the new functional relationships developed. This allowed for pedon SOC and N stores to be calculated for 609 and 468 Alaska pedons respectively, the most available to-date. Additionally, functional relationships were developed for data within soil orders to estimate total SOCHTC and N stores in pedons with missing surface organic horizons where only thicknesses were known. These relationships are presented in order to fill-in missing data and to better define the existing data set for future use. Some 1904 missing Db data points and 1612 corrected SOCHTC data points were added to the total of 4240 points in the 609 pedons that constitute the updated dataset. When O-layer thickness functions developed here were used, SOC and N stores were calculated for an additional 137 and 184 pedons respectively.
