Nooshin Ahmadibaseri; A.A. Sabziparvar; M. Khodamoradpour; L. Alados Arboledas
Abstract
Introduction: Surface Solar Radiation (SSR) as the largest source of land-surface energy is an important parameter in meteorological and climatological studies. Limitations in ground-based measurements have encouraged the users to approach low cost and reliable methods to estimate radiation components, ...
Read More
Introduction: Surface Solar Radiation (SSR) as the largest source of land-surface energy is an important parameter in meteorological and climatological studies. Limitations in ground-based measurements have encouraged the users to approach low cost and reliable methods to estimate radiation components, for the regions where the ground-based radiation data are sparse. Different methods have been developed for estimating SSR including empirical models, radiative transfer models, semi-empirical models, and models based on satellite and reanalysis products. In most studies in Iran, empirical methods have been investigated. Despite the simplicity of these models, they do not accurately represent SSR variations because of not considering all the parameters affecting radiation variations, at large spatial scales with different climates. The Global Land Data Assimilation System (GLDAS) is a combination of measured and satellite data that uses advanced land surface modeling and data assimilation methods. One of the strengths of this model that makes GLDAS unique is that it has global coverage, high spatial-temporal resolution and is available for free. GLDAS is a terrestrial modeling system uncoupled to the atmosphere. This work was aimed to evaluate SSR derived from GLDAS using ground measurements over Iran from 2012 to 2015 on a daily basis.
Materials and Methods: In this study, measured SSR in 24 radiometer stations of Iran from 2012 to 2015 was extracted. Since the measured data are associated with some errors, the quality of the data must be checked and screened before use. In this study, Moradi's proposed method was used to control data quality. The studied areas were classified into three zones of coastal, arid and semi-arid climates based on Digital Elevation Model (DEM) and UNESCO climate classification approach. The GLDAS SSR outputs were extracted with a spatial and temporal resolution of 0.25° grid cell and 3-hourly from 2012 to 2015. The GLDAS is one of the LDAS projects and has been extended jointly by the National Aeronautics and Space Administration (NASA) Goddard Space Flight Center (GSFC) and the National Oceanic and Atmospheric Administration (NOAA) National Centers for Environmental Prediction (NCEP). The purpose of GLDAS is to produce high quality temporal and spatial land surface data. GLDAS drives three land surface models: Mosaic, Noah, and CLM. GLDAS assessments SSR at the land surface using a method and cloud and snow products from the Air Force Weather Agency's (AFWA) Agricultural Meteorology modeling system (AGRMET). Since the GLDAS data are created using the gridded Binary format, the nearest neighborhood interpolation method was used to match these data with ground-based data and GLDAS datasets were generated for station points using CDO software. In this study, GLDAS datasets were compared against measured SSR datasets by four validation metrics. The metrics used are determination coefficient (R2), the mean bias error (MBD), the mean absolute error (MABD), relative mean absolute error (RMABD) and root mean squared error (RMSE).
Results and Discussion: Statistical analysis showed that the performance of GLDAS in SSR evaluation is reasonable in Iran with a high-efficiency coefficient of 0.88. Also, it was shown that the GLDAS has a higher ability to estimate SSR under clear sky (warm seasons) conditions than cloudy conditions (cold seasons). Similar to the obtained results, Träger-Chatterjee et al. (2010); Jia et al. (2013); Boilley and Wild (2015) and Heidary Beni and Yazdanpanah (2017) also showed that the ERA- Interim, NCEP-DOE, RegCM4 and angstrom model are also more capable of estimating SSR in warm seasons. Seasonal bias variations at three studied areas showed that the most changes occurred in summer and least changes in winter. The highest overestimation was also observed in the coastal areas in summer and the lowest overestimation in the semi-arid regions in spring. The evaluation of the GLDAS performance against the site measured SSR data suggests that the GLDAS tends to underestimate in 71% of the studied stations. Moreover, the stations located in the arid region provided a better estimation of SSR as compared with semi-arid and coastal locations. These results were compared with those of Boilley and Wald (2015) that showed ERA-Interim and MERRA reanalysis models have high uncertainty in areas with tropical humid climates, and in regions with arid climates, models perform better in SSR estimation. Our findings were also in good agreement with their results.
Conclusion: GLDAS SSR outputs can be used for agricultural studies. This is due to the facts that arid and semi-arid climates are dominant in Iran and the growing season is mostly in the warm season.
Nooshin Ahmadibaseri; A. Shirvani; mohammad jafar nazemosadat
Abstract
In this study, the artificial neural networks (ANNs) and regression models were used to downscale the simulated outputs of the general circulation models (GCMs). The simulated precipitation for 25.18 º N to 34.51 º N and 45 º E to 60 º E, geopotential height at 850 mb and zonal wind at 200 mb for ...
Read More
In this study, the artificial neural networks (ANNs) and regression models were used to downscale the simulated outputs of the general circulation models (GCMs). The simulated precipitation for 25.18 º N to 34.51 º N and 45 º E to 60 º E, geopotential height at 850 mb and zonal wind at 200 mb for 12.56° N to 43.25° N and 19.68° E to 61.87° E data sets as the predictors were extracted from ECHAM5 GCM for the period 1960-2005. The observed monthly precipitation data of Abadan, Abadeh, Ahwaz, Bandar Abbas, Bushehr, Shiraz and Fasa stations as the predictand were extracted for the period 1960-2005. The principal components (PCs) of the simulated data sets were extracted and then six PCs were considered as the input file of the ANN and multiple regression models. Also the combinations of the simulated data sets were used as the input file of these models. The periods 1960-2000 and 2001-2005 were considered as the train and test data in the ANN, respectively. The Pearson correlation coefficient and normalized root mean square error results indicated that ANN predicts precipitation more accurate than multiple regression. For the monthly time scale, the geopotential height is the best predictor and for the seasonal time scale (winter) the simulated precipitation is the best predictor in ANN based standardized precipitation principal components.
