Construct Atmospheric Humidity Vertical Profile With High Spatial And Time Resolution By High Resolution Mesoscale Numerical Model And Ground-based GPS Data

To forecast the strong convection weather, one of the key problems is to find out theatmospheric vertical stability and wind shear. Now for forecasters, checking the radiosondeobservation is the usual way to obtain information about the atmospheric vertical stability andwind shear. But the spatial and time resolution of the present radiosonde network cannotprovide sufficiently detailed information to indicate the possible appearance of the strongconvective weather. On the other hand, the Mesoscale Numerical Model is able to output allkinds of model variables with high resolution, and furthermore, provide the atmospherictemperature and humidity vertical profiles with high spatial and time resolution. The keyproblem is whether or these profiles are reliable .Using the high resolution radiosonde data of Beijing in June , July, and August , 2004, themesoscale numerical model MM5 with a resolution of 3km of Beijing meteorological bureau ischecked to show its diagnosis ability for the possible occurrence of the convective weather insummer of Beijing, the reliability of the atmospheric temperature and humidity vertical profilesobtained from the numerical model as well as the convection diagnosis parameters which arecalculated from these profiles are estimated. Results show: in all kinds of atmospheric profilesobtained from the numerical model, the wind shear and temperature profiles both are valuableto some extent,while at the location where the profiles show a transition such as an temperatureinversion or a wind direction transition the simulation results are much poorer. The dew pointprofile show large error which cannot reflect the real distribution of the humidity field. Thevertical wind shear of the forecasting deep layer (from ground to 500hpa) shows a relative goodconsistence with the radiosonde observation, while the calculated CAPE value have largedifference with the real value because of the imperfect forecasting results of the temperature andhumidity . The CAPE value must be properly revised in order to diagnose the potential forecastof the strong convection weather. This paper make use of the integrated precipitation water vapor of radiosonde andtomography vapor data of GPS to modify the humidity vertical profile and CAPE valueobtained from the model. Using the integrated precipitation water (PW) of radiosonde, theiterative PW data of radiosonde, and tomography vapor data of GPS respectively, we carry onthree numerical tests. Results show: The efficiency of these three methods is ascending in turn.For the method directly using the radiosonde PW, the efficiency of the forecasting CAPE iseven worse after being modified, which is because that the forecasting atmospheric humiditydistribution is imperfect so that only by modifying the model's PW but not changing thehumidity distribution, the vapor may be larger than the real value and furthermore, and makethe CAPE value larger exceptionally. For the method using the iterative way to carry out thePW modification, due to carry on the modification layer by layer and make a limit for saturationspecific humidity (qs) by the forecasting results so that the humidity cannot increaseexceptionally, the CAPE value becomes much better than the direct method, while its variabletrend is consistence with the observation value and the error decreases a lot, but the forecastingnumerical CAPE value still does not has sufficient accuracy. Finally, directly using thetomography vapor density data of GPS to modify the model, the modified humidity profilebecomes further closer to the observation and the modified CAPE value becomes closer to theobservation either, which indicate that the effect of the modification for the CAPE value isbetter than the previous two methods.