scholarly journals Performance assessment of six bias correction methods using observed and RCM data at upper Awash basin, Oromia, Ethiopia

2021 ◽  
Author(s):  
Bekan Chelkeba Tumsa
Keyword(s):  
Time Series ◽  
Bias Correction ◽  
Swat Model ◽  
Correction Method ◽  
Linear Scaling ◽  
Power Transformation ◽  
Highly Qualified ◽  
Scaling Methods ◽  
Change Impact ◽  
Akaki River

Abstract Selecting a suitable bias correction method is important to provide reliable inputs for evaluation of climate change impact. Their influence was studied by comparing three discharge outputs from the SWAT model. The result after calibration with original RCM indicate that the raw RCM are heavily biased, and lead to streamflow simulation with large biases (NSE = 0.1, R2 = 0.53, MAE = 5.91 mm/°C, and PBIAS = 0.51). Power transformation and linear scaling methods performed best in correcting the frequency-based indices, while the LS method performed best in terms of the time series-based indices (NSE = 0.87, R2 = 0.78, MAE = 3.14 mm/°C, PBIAS = 0.24) during calibration. Meanwhile, daily translation was underestimating simulated streamflow compared with observed and considered as the least performing method. Precipitation correction method has higher visual influence than temperature, and its performance in streamflow simulations was consistent and significantly considerable. Power transformation and variance scaling showed highly qualified performance compared to others with indicated time series value (NSE = 0.92, R2 = 0.88, MAE = 1.58 mm/°C and PBIAS = 0.12) during calibration and validation of streamflow. Hence, PT and VARI methods were the dominant methods which remove biasness from RCM models at Akaki River basin.

scholarly journals Converting Climate Change Gridded Daily Rainfall to Station Daily Rainfall—A Case Study at Zengwen Reservoir

Water ◽  
2021 ◽  
Vol 13 (11) ◽  
pp. 1516
Author(s):  
Tse-Yu Teng ◽  
Tzu-Ming Liu ◽  
Yu-Shiang Tung ◽  
Ke-Sheng Cheng
Keyword(s):  
Climate Change ◽  
Water Resources ◽  
Bias Correction ◽  
Climate Change Impact ◽  
Daily Rainfall ◽  
Correction Method ◽  
Two Stage ◽  
Climate Change Impact Assessment ◽  
Daily Data ◽  
Change Impact

With improvements in data quality and technology, the statistical downscaling data of General Circulation Models (GCMs) for climate change impact assessment have been refined from monthly data to daily data, which has greatly promoted the data application level. However, there are differences between GCM downscaling daily data and rainfall station data. If GCM data are directly used for hydrology and water resources assessment, the differences in total amount and rainfall intensity will be revealed and may affect the estimates of the total amount of water resources and water supply capacity. This research proposes a two-stage bias correction method for GCM data and establishes a mechanism for converting grid data to station data. Five GCMs were selected from 33 GCMs, which were ranked by rainfall simulation performance from a baseline period in Taiwan. The watershed of the Zengwen Reservoir in southern Taiwan was selected as the study area for comparison of the three different bias correction methods. The results reveal that the method with the wet-day threshold optimized by objective function with observation rainfall wet days had the best result. Error was greatly reduced in the hydrology model simulation with two-stage bias correction. The results show that the two-stage bias correction method proposed in this study can be used as an advanced method of data pre-processing in climate change impact assessment, which could improve the quality and broaden the extent of GCM daily data. Additionally, GCM ranking can be used by researchers in climate change assessment to understand the suitability of each GCM in Taiwan.

Modelling runoff generation of a small catchment in the context of climate change by using an ensemble of different climate model outputs and bias correction methods

2020 ◽  
Author(s):  
Kai Sonntag ◽  
Matthias Gassmann
Keyword(s):  
Climate Change ◽  
Bias Correction ◽  
Climate Models ◽  
Climate Model ◽  
Potential Evapotranspiration ◽  
Linear Scaling ◽  
Power Transformation ◽  
Quantile Mapping ◽  
The Future ◽  
Distribution Mapping

<p>Due to climate change, meteorological extremes affect the environment and our society in the past decades. But not only the extremes are piling up, the average temperatures and the precipitation regimes have changed in recent decades. The change in meteorological conditions also affects the water balance and thus also the generation processes of runoff. The aim of this work is to estimate this future change for a small low-mountain catchment in central Germany using climate projections and hydrological modelling.</p><p>As input to the hydrological model HBV Light, climate data from seven different combinations of global and regional climate models are used. However, due to their substantial bias it is necessary to apply bias correction. For each of the three climate input time series used by HBV Light, different bias correction methods are tested: Precipitation (Linear Scaling Multiplication, Quantile Mapping, Power Transformation, Distribution Mapping Gamma), Temperature (Linear Scaling Addition, Quantile Mapping, Variance Scaling, Distribution Mapping Normal) and Potential Evapotranspiration (Linear Scaling Multiplication, Linear Scaling Addition, Quantile Mapping). The corrected climate model outputs are compared to the observed timeseries and rated based on three different efficiency criteria. Overall, the combination of different climate models and bias correction methods generates 63 future hydrological projections. Based on this ensemble, the future water balance of the catchment is assessed. The results show that (1) the biggest uncertainties in the hydrological simulation were generated by uncorrected climate model outputs; (2) the uncertainties in hydrological simulations increase till the end of the century; (3) Power Transformation and Quantile Mapping perform best for precipitation, Linear Scaling Addition and Quantile Mapping for temperature, Linear Scaling Addition and Quantile Mapping for potential evapotranspiration; (4) the total annual outflow increases till 2070 because of an increase of the outflow in winter and spring; (5) in the future, interflow will increase in spring and winter and reduce in summer and autumn; (6) till the end of the century the baseflow will rise in spring and in the rest of year the baseflow will decrease. This study shows that even if changes in the annual total discharge for small catchments have no significant trend, the generation processes and the seasonal values may change in the future.</p>

scholarly journals Addressing the assumption of stationarityin statistical bias correction of temperature

2016 ◽  
Author(s):  
Manolis G. Grillakis ◽  
Aristeidis G. Koutroulis ◽  
Ioannis N. Daliakopoulos ◽  
Ioannis K. Tsanis
Keyword(s):  
Time Series ◽  
Bias Correction ◽  
Statistical Characteristics ◽  
Statistical Bias ◽  
Climate Variable ◽  
Processing Procedure ◽  
Change Impact ◽  
The Mean ◽  
Temperature Time

Abstract. Bias correction of climate variables has become a standard practice in Climate Change Impact (CCI) studies. While various methodologies have been developed, their majority assumes that the statistical characteristics of the biases between the modeled data and the observations remain unchanged in time. However, it is well known that this assumption of stationarity cannot stand in the context of a climate. Here, a method to overcome the assumption of stationarity and its drawbacks is presented. The method is presented as a pre-post processing procedure that can potentially be applied with different bias correction methods. The methodology separates the stationary and the non-stationary components of a time series, in order to adjust the biases only for the former and preserve intact the signal of the later. The results show that the adoption of this method prevents the distortion and allows for the preservation of the originally modeled long-term signal in the mean, the standard deviation, but also the higher and lower percentiles of the climate variable. Daily temperature time series obtained from five Euro CORDEX RCM models are used to illustrate the improvements of this method.

scholarly journals Bias correction method for climate change impact assessments in the Philippines

2013 ◽  
Vol 69 (4) ◽  
pp. I_19-I_24 ◽  
Author(s):  
Cho Thanda NYUNT ◽  
Toshio KOIKE ◽  
Pactricia Ann J. SANCHEZ ◽  
Akio YAMAMOTO ◽  
Toshihoro NEMOTO ◽  
...  
Keyword(s):  
Climate Change ◽  
Bias Correction ◽  
Climate Change Impact ◽  
Correction Method ◽  
The Philippines ◽  
Bias Correction Method ◽  
Change Impact

scholarly journals Selection of Effective GCM Bias Correction Methods and Evaluation of Hydrological Response under Future Climate Scenarios

Climate ◽  
2020 ◽  
Vol 8 (10) ◽  
pp. 108
Author(s):  
Yaogeng Tan ◽  
Sandra M. Guzman ◽  
Zengchuan Dong ◽  
Liang Tan
Keyword(s):  
Climate Change ◽  
Water Resources ◽  
Bias Correction ◽  
Global Climate ◽  
Global Climate Model ◽  
Swat Model ◽  
Correction Method ◽  
Climate Change Scenarios ◽  
Hydrological Variables ◽  
Precipitation And Temperature

Global climate change is presenting a variety of challenges to hydrology and water resources because it strongly affects the hydrologic cycle, runoff, and water supply and demand. In this study, we assessed the effects of climate change scenarios on hydrological variables (i.e., evapotranspiration and runoff) by linking the outputs from the global climate model (GCM) with the Soil and Water Assessment Tool (SWAT) for a case study in the Lijiang River Basin, China. We selected a variety of bias correction methods and their combinations to correct the lower resolution GCM outputs of both precipitation and temperature. Then, the SWAT model was calibrated and validated using the observed flow data and corrected historical GCM with the optimal correction method selected. Hydrological variables were simulated using the SWAT model under emission scenarios RCP2.6, RCP4.5, and RCP8.5. The results demonstrated that correcting methods have a positive effect on both daily precipitation and temperature, and a hybrid method of bias correction contributes to increased performance in most cases and scenarios. Based on the bias corrected scenarios, precipitation annual average, temperature, and evapotranspiration will increase. In the case of precipitation and runoff, projection scenarios show an increase compared with the historical trends, and the monthly distribution of precipitation, evapotranspiration, and runoff shows an uneven distribution compared with baseline. This study provides an insight on how to choose a proper GCM and bias correction method and a helpful guide for local water resources management.

scholarly journals A Quality-Control and Bias-Correction Method Developed for Irregularly Spaced Time Series of Observational Pressure Data

2011 ◽  
Vol 28 (10) ◽  
pp. 1317-1323 ◽  
Author(s):  
Stefan Sperka ◽  
Reinhold Steinacker
Keyword(s):  
Time Series ◽  
Bias Correction ◽  
Potential Temperature ◽  
Slight Modification ◽  
Correction Method ◽  
Mean Deviation ◽  
Homogeneity Test ◽  
Standard Normal Homogeneity Test ◽  
Functional Fitting ◽  
Break Points

Abstract This paper presents a method to detect and correct occurring biases in observational mean sea level pressure (MSLP) data, which was developed within the Mesoscale Alpine Climate Dataset [MESOCLIM; i.e., 3-hourly MSLP, potential and equivalent potential temperature Vienna Enhanced Resolution Analysis (VERA) analyses for a 3000 km × 3000 km area centered over the Alps during 1971–2005] project. There are many reasons for a change of a measurement site’s performance, for example, a change in the instrumentation, a slight modification of the site’s place or position, or a different way of data processing (pressure reduction). To get an estimate for these artificial influences in the data, deviations for each reporting station at each point of time were calculated, using a piecewise functional fitting approach that is based on a variational algorithm. In this algorithm first- and second-order spatial derivatives are minimized using the tested stations neighbor stations and furthermore their neighbors. The resulting time series of deviations for each station were then tested with a “standard normal homogeneity test” to detect changes in the mean deviation. With the knowledge of these “break points,” bias-correction estimates for each station were calculated. These correction estimates are constant between the detected break points because the method does not detect different slopes in trends. Application of these correction estimates yields in smoother fields and a more homogenous distribution of trends.

scholarly journals Impact of Bias Correction on Present and Mid-Future Rainfall Projections over Senegal

2021 ◽  
Vol 17 (37) ◽  
pp. 137
Author(s):  
Sarr Alioune Badara ◽  
Diatta Samo ◽  
Kébé Ibourahima ◽  
Sultan Benjamin ◽  
Camara Moctar
Keyword(s):  
Bias Correction ◽  
Regional Climate ◽  
Correction Method ◽  
Climate Change Signal ◽  
Linear Scaling ◽  
Rainfall Events ◽  
Mean Intensity ◽  
Bias Correction Method ◽  
The Mean ◽  
The Impact

In this study, we analyze the impact of bias correction models on present and future precipitation and extremes rainfall events over Senegal. The commonly used linear scaling (LS) bias correction method has been applied on four (4) regional climate models (RCMs) of the Coordinated Regional Climate Downscaling Experiment (CORDEX) program. The linear scaling bias correction method was firstly calibrated and validated during the 1976-1990 and 1991-2005 periods, respectively. The comparison with the observed data revealed that the linear scaling method significantly improves the mean and the extreme precipitations during the validation period. The RCMs generally simulate a decrease of rainfall in the mid-twenty-first century under the RCP8.5 greenhouse gas concentration pathway compared to the reference period (1976-2005), except for the CCLM4 and the RCA4 models which show respectively a slight increase overall Senegal and the east of the country. The changes in precipitation indices such as the number of wet days (R1mm) and mean frequency of heavy rainfall events (R20mm) follows that mean precipitation change distribution. Almost uncorrected RCMs (except RCA4) predict during the near future an increase in of the mean intensity of daily rainfall events (SDII), the mean intensity of precipitation events above the 95th Percentile (R95PTOT) and the mean maximum dry spells length (CDD), whereas a decrease in the mean maximum wet spells length (CWD) is projected. After applying the LS bias correction, the spatial distribution patterns are not so much modified in all the models but the magnitude of the climate change signal is either amplified or moderated depending on the considered variables.

scholarly journals Random Forest Regression Model for Estimation of the Growing Stock Volumes in Georgia, USA, Using Dense Landsat Time Series and FIA Dataset

2021 ◽  
Vol 13 (2) ◽  
pp. 218
Author(s):  
Shingo Obata ◽  
Chris J. Cieszewski ◽  
Roger C. Lowe III ◽  
Pete Bettinger
Keyword(s):  
Time Series ◽  
Random Forest ◽  
Land Cover ◽  
Satellite Imagery ◽  
Bias Correction ◽  
Forest Disturbance ◽  
Correction Method ◽  
Volume Estimation ◽  
Series Data ◽  
Bias Correction Method

The forest volumes are essential as they are directly related to the economic and environmental values of the forests. Satellite-based forest volume estimation was first developed in the 1990s, and the accuracy of the estimation has been improved over time. One of the satellite-based forest volume estimation issues is that it tends to overestimate the large volume class and underestimate the small volume class. Free availability of the major satellite imagery and the development of cloud-based computational platforms facilitate an immense amount of satellite imagery in the estimation. In this paper, we set three objectives: (1) to examine whether the long Landsat time series contributes to the improvement of the estimation accuracy, (2) to explore the effectiveness of forest disturbance record and land cover data as ancillary spatial data on the accuracy of the estimation, and (3) to apply the bias correction method to reduce the bias of the estimation. We computed three Tasseled-cap components from the Landsat data for preparation of short (2014–2016) and long (1984–2016) time series. Each data entity was analyzed with harmonic regressions resulting in the coefficients and the fitted values recorded as pixel values in a multilayer raster database. Data included Forest Inventory and Analysis (FIA) unit field inventory measurements provided by the United States Department of Agriculture Forest Service and the National Land Cover Database and disturbance history data added as ancillary information. The totality of the available data was organized into seven distinct Random Forest (RF) models with different variables compared against each other to identify the ones with the most satisfactory performance. A bias correction method was then applied to all the RF models to examine the effectiveness of the method. Among the seven models, the worst one used the coefficients and fitted values of the short Landsat time series only, and the best one used coefficients and fitted values of both short and long Landsat time series. Using the Out-of-bag (OOB) score, the best model was found to be 34.4% better than the worst one. The model that used only the long time series data had almost the same OOB score as the best model. The results indicate that the use of the long Landsat time series improves model performance. Contrary to the previous research employing forest disturbance data as a feature variable had almost no effect on OOB. The bias correction method reduced the relative size of the bias in the estimates of the best model from 3.79% to −1.47%, the bottom 10% bias by 12.5 points, and the top 10% bias by 9.9 points. Depending on the types of forest, important feature variables were differed, reflecting the relationship between the time series remote sensing data we computed for this research and the forests’ phenological characteristics. The availability of Light Detection And Ranging (LiDAR) data and accessibility of the precise locations of the FIA data are likely to improve the model estimates further.

Sign in / Sign up

Export Citation Format

Share Document