Application Of Satellite Retrieval Data Of Surface Solar Radiation In Building Heating Research For Xinjiang

Xinjiang has abundant solar resources throughout the year,but there are only 11 ground radiation observation stations that record radiation-related parameters.While satellite retrieval data can make up for the lack of ground observation data.Therefore,understanding the satellite retrieval error of the surface solar radiation and using the satellite retrieval radiation to simulate building energy consumption can provide a reference for the application of satellite data in building heating and development an utilization of solar energy.Therefore,Aerosol optical depth（AOD）derived from Aqua MODIS C6.1 dataset and cloud optical depth（COD）derived from CERES SSF Aqua MODIS Edition 4A dataset were used to study the temporal and spatial distribution of AOD and COD in 2003-2018 firstly.And then taking the Kashi site as an example,combined with the Accu RT radiative transfer model,the major impact of aerosol and cloud on the solar radiation was quantified;Then surface solar radiation data from ground sites and satellite inversions were used to study temporal distribution,spatial distribution of annual average,heating and non-heating period in 2017.And the best fitting result of radiation from ground observations and satellite inversions were explored during the year and the heating period under six sky conditions:cloudless clean sky,cloudless unclean sky,few cloud clean sky,few cloud unclean sky,and satellite retrieval error was analyzed briefly.Finally,taking Qitai County in severe cold area as a research object,the temperature fields,velocity fields of the room were simulated and the heat transfer and heat transfer coefficient of the outer protective structure were calculated.The main conclusions are as follows:（1）The annual variation of AOD in Xinjiang showed a bi-peak distribution,and the average AOD during non-heating period was higher than that during heating period;The grid average AOD was between 0.0180–1.6703;For the heating period,the high–value areas of AOD in southern Xinjiang were mainly distributed around the Tarim Basin,while the high-value areas of northern Xinjiang were mainly distributed in the Junggar Basin;for non-heating periods,the high-value areas of AOD in southern Xinjiang were Tarim Basin.The grid average value AOD in the heating period in Xinjiang was between 0.05–0.10,and the maximum of distribution frequency was at 36°N,which was 60%;the grid average AOD in the non-heating period is generally less than 0.45,and the maximum of distribution frequency was at 36°N,which was 82%（2）The annual variation of COD in Xinjiang showed a bi-peak distribution,and the average value COD during non-heating period was higher than that during heating period;The multi-year average COD in Xinjiang was between 7.13–7.82.The multi-year grid average COD in Xinjiang was between 5.23–11.00,and the distribution characteristics of the grid average COD in the heating period in northern Xinjiang were similar to that of many years;the grid average COD in the non-heating period was significantly higher than that of multi-year average and in heating period.The frequency of grid average COD within 5.0 in northern Xinjiang was mostly above22%.The frequency of the grid average COD was most in the range of less than 2.5;the frequency of the grid average COD between 2.5-5.0 was the maximum in southern Xinjiang,which was28%.（3）Taking the Kashi site as an example,combined with the Accu RT radiative transfer model,the impact of AOD,COD on the solar radiation was analyzed.When AOD increased from 0.05 to0.80,the total radiation in the heating+non-heating period,heating period and non-heating period change from 733.10 W·m^-2 to 417.78 W·m^-2,734.73 W·m^-2 to 412.70 W·m^-2,and 731.52 W·m^-2 to423.10 W·m^-2 respectively.When COD of water cloud increased from 5.00 to 13.00,SFD decreased by 166–170 W·m^-2;When cod of ice cloud increased from 4.00 to 16.00,the total radiation decreased by 229–235 W·m^-2,and the scattering radiation was basically consistent with the total radiation.The sensitivity of the total downward shortwave radiation and scattered radiation caused by the decrease of AOD or COD was significantly higher than that caused by the increase of AOD or COD.（4）Except Ruoqiang site,the annual variation of the surface solar radiation at the other 10sites showed a single peak distribution;The average value of The surface solar radiation in non-heating period at the Hami station was the maximum value,the smallest value in heating period at Urumqi,which was 322.77 W·m^-2 and 70.69 W·m^-2,respectively;The annual grid average value of the surface solar radiation was concentrated at 211?373 W·m^-2,and the grid average radiation in the heating period and non-heating period were distributed at 132?341 W·m^-2and 242?444 W·m^-2 in Xinjiang in 2017,respectively.The grid average radiation in southern Xinjiang are generally higher than those in northern Xinjiang.（5）In 2017,when there were clear and cloudless sky during the whole year and heating period,and non-clear and partly cloudy throughout the year,the satellite retrieval accuracy was the highest at Yining and Aksu sites;When there were non-clear and cloudless sky throughout the year,the satellite retrieval accuracy were the best at Yanqi and Altay sites.The satellite retrieval effect at the Altay site was the best on the non-clear and cloudless sky and non-clear and partly cloudy sky during the heating period.On clear and partly cloudy sky,compared with clear and cloudless,the dispersion of matching surface solar radiation of ground observation results and satellite retrieval data was significantly higher than that on clean and cloudless,and the retrieval effect was significantly worse than the latter.On mostly cloudy sky and overcast sky,the satellite retrieval solar radiation was significantly affected by cloud,and there existed different degrees of overestimation,and the satellite retrieval effects at 11 sites were all biased.Therefore,on mostly cloudy days,RMSE of 11 sites exceeded 900,even if RB of some sites were small,and R² was basically less than 0.75.On overcast sky,RMSE of 11 sites was basically more than 800,and only RMSE of Urumqi station was less than 800 in heating period,but RB of this station was 71.17%.（6）Taking the Qitai county as an example,room with glass curtain wall as the numerical simulation mode was established,where heating mode was air conditioner for up-inlet and up-outlet.When the ground observation radiation value of Urumqi is taken as the input data of CFD,the heat gain of the room heated by solar radiation is less,and the room heating load is large;the indoor air temperature is basically 291 K（18℃）;and on the 1.1 m plane where the working face of human body is located,the glass curtain wall is greatly affected by the external cold air.When the radiation value retrieved from Qitai County is used as the CFD input,the room air temperature is slightly higher than that when the ground observation radiation value of Urumqi is used as the simulation input,and the working face position temperature is basically maintained at292 K（19℃）,which can barely meet the heating design requirements.The results show that the heating energy consumption of buildings is lower than that of Urumqi ground observation when the satellite retrieved radiation value is used as input.The greater the solar radiation is,the smaller the heating energy consumption is.For Qitai County without ground data,the deviation of average building heating energy consumption and average heating index was up to 33.96%when the ground radiation data of adjacent cities was used as the simulation value.So correcting the satellite retrieval solar radiation data can provide a reference for correcting building heating energy consumption and for the application of satellite data in building heating.