| 摘要: | 本研究目標在於將目前台灣地區常用於模擬及預報之 Goddard Cumulus
Ensemble(GCE)微物理參數化方案，從原先的單矩量形式，轉變為自由度更高，且提供更多水物粒子相關資訊之雙矩量形式。此次更新完之微物理參數化方案共預報七種不同水物粒子(水氣、雲水、雨水、雲冰、雪、軟雹與冰雹)，以及兩種暖雨過程之總水物粒子濃度(雲水與雨水)。為了檢視雙矩量之 GCE 微物理參數化方案(GCEDM)在模式中的降雨預報表現，故使用 1) 2D Thunderstorm 理想化實驗、2) 3D Supercel 理想化實驗
以及 3) 暖雨過程主導之台灣降水真實個案，並主要以後兩者來與不同之微物理參數化方案進行評估與比較。
2D Thunderstorm 理想化實驗初步顯示了 GCESM 與 GCEDM 在降水能力與分布的相似與差異之處。而在 3D Supercel 理想化實驗中，GCEDM 所得到之累積降雨量相較於單矩量之 GCE 微物理參數化方案(GCESM)有所減少，但在動力、熱力及降雨率等特徵兩者依然相似。此外，當增加了總水物粒子濃度的計算後，GCEDM 在對流系統與層狀系統中，能夠像其他雙矩量微物理方案一樣，呈現出更多樣化的雨滴粒徑分布，使其更接近真實情形。另一方面，在真實個案之定量降水預報上，GCEDM 整體表現出了比 GCESM 更好的降雨預報能力，特別是在強降雨區域發生暖雨過程主導之降水時有明顯改善。原因在於透過增加雨水粒子濃度預報，微物理過程中作為主要雨水生成作用之雲雨水聚集作用有所減弱，大幅減緩了原先 GCESM 中過報的累積降雨情形。在動力方面，GCEDM 之低層輻合作用較 GCESM 有所減弱，底層水氣難以透過上升運動到達高層，故高層飽和度較低，水物粒子轉化上也相較困難，導致降雨量上的大幅減少。此預報結果顯示出 GCEDM 不只能在實際個案中提供更詳盡的水物粒子資訊，甚至可以提供相比 GCESM 更好的預報效益。;This study aims to modify Goddard Cumulus Ensemble microphysics (GCE) scheme,which is the widely used microphysics scheme for simulations and forecasts around the world, from single-moment to double-moment scheme. The upgraded scheme predicts the mixingratios of seven species hydrometeors (water vapor, cloud water, rain water, cloud ice, snow,graupel and hail), and total number concentrations of hydrometeors in warm-rain processes(cloud water and rain water). To examine the performance of GCE double-moment, it was evaluated and compared to different microphysics schemes by: 1) an idealized 2D thunderstorm, 2) an idealized 3D supercell storm tests and 3) a warm-rain processes dominant real case over Taiwan.
Similarities and differences of precipitation between GCESM and GCEDM examine by 2D idealized thunderstorm test. For the 3D supercell idealized test, the result of accumulated rainfall is less in GCE double-moment compared to GCE single-moment, but the features of the dynamic, thermodynamic, and rainfall rate are similar as GCE single-moment scheme. In addition, when calculating the total number concentration, GCE double-moment scheme shows the capability to present more diversity of rain droplet sizes in both convective and stratiform regions toward reality, and this is similar as other double-moment schemes. On the other hand, the performance of quantitative precipitation forecast showed that, the GCE double-moment scheme is had better forecast skill compared to the GCE single-moment, especially for the warm-rain processes dominant situation in heavy rainfall region. By appending rain number concentration forecast, accretion of cloud and rain as main rainwater source microphysics process becomes weaker. It leads to reduce overestimate of accumulated rainfall obviously. For dynamic, low level convergence in GCEDM is weaker than GCESM. Low level water vapor is hard to arrive high level by updraft. Therefore, lower saturation and more difficult to transform into hydrometeors. Finally, it makes precipitation decrease dramatically. In summary, GCEDM
not only provides more detail hydrometeor particle information, but also can get better benefit of forecast. |
