The Applicability of NCEP Reanalysis Data to Severe Convection Environment Analysis

Operational sounding observation is carried out twice per day, but four similar vertical profiles can be obtained from NCEP reanalysis data. Therefore, the vertical profiles obtained from NCEP reanalysis data are assessed and its applicability in diagnosing severe convection environment is analyzed. Sixty soundings in close proximity to supercell storms are investigated. These supercell storms are observed in China and identified by Doppler weather radar. The soundings which may be polluted by storm are replaced by soundings from its downstream area or upstream area. The profiles obtained from NCEP data are at the same time and as close to the observation soundings as possible. The vertical profile data that obtained from NCEP reanalysis data are compared with soundings. The ingredients-based methodology is used to discuss the potential for severe convection. Vertical wind shear is the first element to check. NCEP wind in middle and high troposphere is almost consistent with observation, so the deep and middle vertical wind shear (0—6 km, 0—8 km and 0—3 km) and other related dynamical parameters can be calculated from NCEP data except for 0—1 km low vertical wind shear, because wind difference is significant in planet boundary layer (PBL). To study instability of storms, convective available potential energy (CAPE), temperature lapse rate (temperature difference between 850 hPa and 500 hPa), and K index are investigated. The statistic results show that the difference of CAPE between observation and NCEP profiles is significant, because CAPE is sensitive to dewpoint and temperature of lifting air mass. On average, 1℃ temperature increment of lifting airmass brings 200 J·kg^-1CAPE augmentation, 1℃ dewpoint increases CAPE by nearly 500 J·kg^-1, and the augmentation can vary from 0 to 1000 J·kg^-1. Moisture is one of the three ingredients for thunderstorm. The error of NCEP moisture parameter is significant, especially within planet boundary layer, the average dewpoint difference between NCEP and observation is 2℃ in low troposphere, which results in nearly 1000 J·kg^-1CAPE difference. To calculate CAPE from NCEP data, moisture should be corrected according to observation. Temperature lapse rate can be used to diagnose atmospheric instability instead of CAPE, as temperature profile can be used to analysis severe convection. The difference of K index is small in most cases. NCEP output variable CAPEsfc (surface CAPE) is unreasonably small, but the tendency can indicate the change of CAPE. Most lifting processes are within PBL, where the difference of the atmospheric parameters especially the wind direction between observation and NCEP data is significant, so it's not suitable to use NCEP data to study lifting mechanism of thunderstorm. On average, the NECP moisture profile is much drier at low level and wetter at middle level than observation, and wind speed above 925 hPa is weaker than observation, which lowers the possibility of severe convection.