AN EMPIRICAL METHOD FOR ESTIMATING SURFACE LONG-WAVE RADIATION EXCHANGES IN THE QINGHAI-XIZANG PLATEAU
Zhou Yun-hua (Institute of Geography, Academia Sinica)
A meteorological experiment was carried out in the Qinghai-Xizang Plateau during May-August, 1979. Six heat source observational points were established to study surface radiation and head budget over the Qinghai-Xizang Plateau (Tab. 1). Surface long-wave radiation were measured ten times a day using Eppley precision infared radiometer (Model RIR). Relationships between surface long-wave radiation and surface meteorological elements were analysed under various sky conditions, which allow the following results: (1) The isothermal emissivity, e(H2O+ CO2), of atmospheric water vapour and carbon dioxide was calculated from staley's flux emissivity tables by monthly mean radiosonde data of twenty aerological stations in the Qinghai-Xizang Plateau and its adjacent areas during May-August, 1979, for the sake of studing relationship between the downward long-wave radiation and the optical depth of water vapour. Results indicate that e (H2O+CO2) varies nearly linearly with the logarithm of the monthly mean effective optical depth of water vapour (W) (Fig. 1), where is acceleration of gravity; P and P0 are atmospheric pressure at surface and standard surface (1000mb); q is specific humidity. (2) The ratio of W to the monthly mean amount of precipitable water vapour W can be related by with correlation coefficient 0.999 (Fig. 2), and W (in cm) can be related to the monthly mean surface vapour pressure e (in mb) by with correlation coefficient 0.991 (Fig. 3) Combination of eqs. (1) and (2) gives logjoW=-1.018 + 0.880 log10 + 1.128log10e (3) (3) The regression of e(H2O + CO2) on P and e is e(H20 + C02) =0.595 + 0.289 Iog10 p/p+ 0.1551og10e (4) The mean relative error of estimated values of e(H2O + CO2) from eqs. (4) to values calculated from Staley's flux emissivity tables is 1.0%. (4) From radiation measurements at six sites, the mean apparent atmospheric emissivity eoj for clear skies, defined as the ratio of down ward long-wave radiation to black-body ra-diation at screen temperature, can be related to P (in mb) and e (in mb) by with mean relative error 2.5%. The mean relative error Aeo for every site was tabulated in Tab. 2. (5) The total downward radiation G, on the surface, under cloudy sk'y, Can been wr-itten as where n is mean fractional cloudiness; oTj is black-body radiation at mean screen temperature Tajk is an empirical coefficient. The approximatly relation between k and e is k= 0.200-0.005e (7) By substituting eq. (7) into eq. (6), the relation becomes with monthly and decadal root-mean-square error 21 and 22 cal/cm2 ?day and mean relative error 2.6% and 2.6% respectivly. The monthly and decadal mean relative errors AG for every site were tabulated in Tab. 3. (6) Net long-wave radiation F can be calculated from where U is monthly or decadal mean upward long-wave radiation on surface; aT40 is black-body radiation at mean surface temperature is an empirical coefficient and equal to 1.016 if U is monthly means based on the 10 times observations per day and T0 is the regular weather station means where the obser vations are taken 4 times per day. The monthly and decadal root-mean-square error of estimated values F is 20 and 22cal/cm2 ?day and mean relative error is 6.9% and 8.0% respectivly. The mean monthly and decadal relative errors AF for every site were tabulated in Tab. 4.