Operational convective-scale data assimilation over Iran: A comparison between WRF and HARMONIE-AROME

The impact of applying three-dimensional variational data assimilation (3D-Var DA) on convective-scale forecasts is investigated by using two mesoscale models, the Weather Research and Forecasting model (WRF-ARW) and the Hirlam and Aladin Research Model On Non-hydrostatic-forecast Inside Europe (HARMONIE-AROME). One month (1 to 30 December 2013) of numerical experiments were conducted with these two models at 2.5 km horizontal resolution, in order to partly resolve convective phenomena, on the same domain over a mountainous area in Iran and neighboring areas. Furthermore, in order to estimate the domain specific background error statistics (BES) in convective scales, two months (1 November to 30 December 2017) of numerical experiments were carried out with both models by downscaling operational ECMWF forecasts. For setting the numerical experiments in an operational scenario, ECMWF operational forecast data were used as initial and lateral boundary conditions (ICs/LBCs). In order to examine the impact of data assimilation, the 3D-Var method in cycling mode was adopted and the forecasts were verified every 6 hours up to 36 hours for selected meteorological variables. In addition, 24 h accumulated precipitation forecasts were verified separately. Generally, the WRF and HARMONIE-AROME exhibit similar verification statistics for the selected forecast variables. The impact of DA on the numerical forecast shows some evidence of improvement in both models, and this effect decreases severely at longer lead times. Results from verifying the 24 h convective-scale precipitation forecasts from both models with and without DA suggest the superiority of the WRF model in forecasting more accurately the occurred precipitation over the simulation domain, even for the downscaling run.     

A comparison of reanalyses in the tropical stratosphere. Part 3: inclusion of the pre-satellite data era

 In this study, previous evaluations of the monthly mean structure of the tropical lower stratosphere in reanalyzed datasets are extended to include the period 1958–1978, when no satellite-based observations were available. It is shown that a large discontinuity, in temperatures near the tropical tropopause, in the NCEP data occurred when the Tiros Operational Vertical Sounder (TOVS) became operational. When only rawinsonde data were available, the tropopause temperatures in the NCEP dataset are in better agreement with ERA data for TOVS period. Both NCEP and NASA reanalyses show similar deviations from the ERA data in the TOVS renalyses show similar deviations from the ERA data in the TOVS period. There is also a stepwise change in the lower stratospheric meridional velocity when the TOVS data were introduced into the NCEP reanalyses. This discontinuity is such that in the 1958–1978 period, the annual cycle in zonal mean meridional velocity in the NCEP data resembles that of the ERA data in the 1979–1993 period. The differences are shown to result from large changes in the local meridional flow in the Indonesian region. The temporal consistency of the QBO is examined; it is shown that the NCEP assimilation system is sensitive to the data available. There is a change in the zonally asymmetric structure of the zonal wind over time, presumably related to the changes in input data and the inability of the model to represent the three-dimensional structure of the tropical lower stratosphere. These results provide further evidence of the value of rawinsonde data in data assimilation systems as well as the need to use satellite radiance data in an appropriate manner. Received: 7 April 1997 / Accepted: 4 September 1998     

Impact of single-point GPS integrated water vapor estimates on short-range WRF model forecasts over southern India

Specifying physically consistent and accurate initial conditions is one of the major challenges of numerical weather prediction (NWP) models. In this study, ground-based global positioning system (GPS) integrated water vapor (IWV) measurements available from the International Global Navigation Satellite Systems (GNSS) Service (IGS) station in Bangalore, India, are used to assess the impact of GPS data on NWP model forecasts over southern India. Two experiments are performed with and without assimilation of GPS-retrieved IWV observations during the Indian winter monsoon period (November–December, 2012) using a four-dimensional variational (4D-Var) data assimilation method. Assimilation of GPS data improved the model IWV analysis as well as the subsequent forecasts. There is a positive impact of ～10 % over Bangalore and nearby regions. The Weather Research and Forecasting (WRF) model-predicted 24-h surface temperature forecasts have also improved when compared with observations. Small but significant improvements were found in the rainfall forecasts compared to control experiments.     

Effect of 3DVAR assimilation of MODIS temperature and humidity profiles on the dynamic and thermodynamic features of three monsoon depressions over the Bay of Bengal

The three-dimensional variational data assimilation (3DVAR) technique in the advanced weather research and forecast model is used to study the impact of assimilating Moderate Resolution Spectroradiometer (MODIS) retrieved temperature and humidity profiles on the dynamic and thermodynamic features for three monsoon depressions over the Bay of Bengal, India. For better understanding of the role of various physical processes in the evolution of monsoon depression, a detailed diagnostic study is performed on all the three depression cases. Numerical experiments were conducted in a system of two-way nested domains with a horizontal resolution of 36 and 12 km, respectively. The assimilation of MODIS data did improve the mean sea level pressure patterns and spatial distribution of rainfall patterns in all the three monsoon depression cases studied. Higher values of equitable threat score and lower bias values are seen consistently for the entire rainfall threshold range and for all the three depression cases with 3DVAR assimilation of MODIS temperature and humidity profiles. The current operational regional models in India do not ingest the MODIS temperature and humidity profiles and hence the present study is particularly relevant to the operational forecasting community in India in their ongoing efforts to improve weather forecasting over India.     

A COMPARISON OF THE BLENDING AND CONSTRAINING METHODS TO INTRODUCE LARGE-SCALE INFORMATION INTO GRAPES MESOSCALE ANALYSIS

To solve the problem of mesoscale analysis error accumulation after a period of continuous cycle data assimilation (CCDA), a blending method and a constraining method are compared to introduce global analysis information into the Global/Regional Assimilation and Prediction Enhanced System mesoscale three-dimensional variational data assimilation system (GRAPES-Meso 3Dvar). Based on a spatial filter used to obtain a blended analysis, the blending method is weighted toward the T639 global analysis for scales larger than the cutoff wavelength of 1,200 km and toward the GRAPES mesoscale analysis for wavelengths below that. The constraining method considers the T639 global analysis data as an extra source of information to be added in the 3DVar cost function. The cloud-resolving GRAPES-Meso system (3 km resolution) with a 3 h analysis cycle update is chosen, and forecast experiments on an extreme precipitation event over the eastern part of China are presented. The comparison shows that the inclusion of large-scale information with both methods has a positive impact on the regional model, in which the 3 h background forecasts are slightly closer to the radiosonde observations. The results also show that both methods are effective in improving large-scale analysis while reserving the well-featured mesoscale information, leading to an enhancement in the balance and accuracy of the analysis. Subjective verification reveals that the introduction of large-scale information has a visible beneficial impact on the forecast of precipitation location and intensity. The methodologies and experiences presented in this paper could serve as a reference for ongoing efforts toward the development of multi-scale analysis in GRAPES-Meso.     

Intercomparison of the performance of MM5/WRF with and without satellite data assimilation in short-range forecast applications over the Indian region

The present study is conducted to verify the short-range forecasts from mesoscale model version5 (MM5)/weather research and forecasting (WRF) model over the Indian region and to examine the impact of assimilation of quick scatterometer (QSCAT) near surface winds, spectral sensor microwave imager (SSM/I) wind speed and total precipitable water (TPW) on the forecasts by these models using their three-dimensional variational (3D-Var) data assimilation scheme for a 1-month period during July 2006. The control (without satellite data assimilation) as well as 3D-Var sensitivity experiments (with assimilating satellite data) using MM5/WRF were made for 48 h starting daily at 0000 UTC July 2006. The control run is analyzed for the intercomparison of MM5/WRF short-range forecasts and is also used as a baseline for assessing the MM5/WRF 3D-Var satellite data sensitivity experiments. As compared to the observation, the MM5 (WRF) control simulations strengthened (weakened) the cross equatorial flow over southern Arabian sea near peninsular India. The forecasts from MM5 and WRF showed a warm and moist bias at lower and upper levels with a cold bias at the middle level, which shows that the convective schemes of these models may be too active during the simulation. The forecast errors in predicted wind, temperature and humidity at different levels are lesser in WRF as compared to MM5, except the temperature prediction at lower level. The rainfall pattern and prediction skill from day 1 and day 2 forecasts by WRF is superior to MM5. The spatial distribution of forecast impact for wind, temperature, and humidity from 1-month assimilation experiments during July 2006 demonstrated that on average, for 24 and 48-h forecasts, the satellite data improved the MM5/WRF initial condition, so that model errors in predicted meteorological fields got reduced. Among the experiments, MM5/WRF wind speed prediction is most benefited from QSCAT surface wind and SSM/I TPW assimilation while temperature and humidity prediction is mostly improved due to latter. The largest improvement in MM5/WRF rainfall prediction is due to the assimilation of SSM/I TPW. The assimilation of SSM/I wind speed alone in MM5/WRF degraded the humidity and rainfall prediction. In summary the assimilation of satellite data showed similar impact on MM5/WRF prediction; largest improvement due to SSM/I TPW and degradation due to SSM/I wind speed.     

Impact of MSMR data on NCMRWF Global Data Assimilation System

Summary ?It is very essential to make use of non-conventional remotely sensed data mainly from various satellites for numerical weather prediction. IRS series of Indian satellites have been found very useful for various types of studies. Recently IRS-P4 was successfully launched and its Multi-frequency Scanning Microwave Radiometer (MSMR) sensor is giving near surface wind speed and total precipitable water content over oceanic region. An attempt has been made to assimilate directly these two geophysical parameters derived at 150 km resolution, along with other global meteorological data, in the National Center for Medium Range Weather Forecasting (NCMRWF), New Delhi Global Data Assimilation System. (GDAS). The paper describes the development work done in utilizing the surface wind speed and the total precipitable water content data in the NCMRWF operational GDAS. The basic algorithms for assimilating the MSMR data directly in the global analysis scheme have been described. The analyzed fields produced after running the six hourly GDAS cycle have been examined and various aspects of the impact of this data set on the global analysis especially for Indian oceanic region have been evaluated. Other aspects like, penalty contributions, root mean square (rms) errors of various types of data both with respect to the analysis and the background field, etc. have been examined. The impact of additional IRS-P4 data on assimilation and model simulation is found to be positive and beneficial. Present affiliation: Department of Meteorology, Florida State University, Tallahassee, Florida 32306, USA Received February 26, 2001; revised February 27, 2002     

Intercomparison of the characteristic features of the southwest monsoon over india as reproduced in the mean monthly analyses and forecasts of some operational NWP centres during 1995

Summary An intercomparison of the characteristic features of the Indian summer monsoon has been carried out for the monsoon months (June to September) of 1995 using the mean monthly analyses/forecasts from the operational centres of ECMWF, JMA, UKMO and NCMRWF. This exercise was undertaken to determine how well the large scale monsoon features over India were reproduced in the operational output in 1995 and also to assess the performance of the NCMRWF assimilation/forecast system. For this purpose, precipitation, mean sea level pressure, circulation features in the lower (850 hPa) and upper (200 hPa) troposphere, mid-tropospheric (500 hPa) temperature, and latent heat flux were examined.It is found that all the dominant features of the Indian summer monsoon are fairly well represented in the analysis and medium range forecasts of the ECMWF, JMA and UKMO. The NCMRWF output agrees well with those from other centres except for a sharp gradient in precipitation across the west coast which was not captured well in the forecasts due to the relatively coarse horizontal resolution of the model compared to that used at other operational centres. Other important features of the southwest monsoon, like the heat low over the northwestern part of the country, the lower level westerly jet and upper level easterly jet etc. are found to be reasonably well represented in the output of all operational centres. The JMA analyses and forecasts possessed greater levels of moisture compared to the NCMRWF output possibly due to the synthetic moisture information used at JMA. The evolution characteristics of the summer monsoon onset over the southern tip of India are found to be comparable in the output of JMA and NCMRWF.With 13 Figures     

The impact of inhomogenous background errors on a global wave data assimilation system

One of the main limitations in current wave data assimilation systems is the lack of an accurate representation of the structure of the background errors. In this work, models for the observational error variance, background error variance and background error correlations are developed based on the results of previous studies. These are tested in a global wave data assimilation system and the resulting wave forecasts are verified against independent observations from buoys. Forecasts of significant wave height show substantial improvement over the Australian Bureau of Meteorology's current operational wave forecasting system. However, forecasts of peak period are not similarly improved. The regional impacts of the new assimilation scheme are found to vary on a seasonal basis. Overall, it is shown that the inclusion of a latitudinally dependent background error, and improved specification of the background and observational error variances can reduce the root-mean-square error of 24-hour forecast Significant Wave Height by almost 10%.     

Simulation of heavy rainfall events over Indian monsoon region using WRF-3DVAR data assimilation system

We present the results of the impact of the 3D variational data assimilation (3DVAR) system within the Weather Research and Forecasting (WRF) model to simulate three heavy rainfall events (25–28 June 2005, 29–31 July 2004, and 7–9 August 2002) over the Indian monsoon region. For each event, two numerical experiments were performed. In the first experiment, namely the control simulation (CNTL), the low-resolution global analyses are used as the initial and boundary conditions of the model. In the second experiment (3DV-ANA), the model integration was carried out by inserting additional observations in the model’s initial conditions using the 3DVAR scheme. The 3DVAR used surface weather stations, buoy, ship, radiosonde/rawinsonde, and satellite (oceanic surface wind, cloud motion wind, and cloud top temperature) observations obtained from the India Meteorological Department (IMD). After the successful inclusion of additional observational data using the 3DVAR data assimilation technique, the resulting reanalysis was able to successfully reproduce the structure of convective organization as well as prominent synoptic features associated with the mid-tropospheric cyclones (MTC). The location and intensity of the MTC were better simulated in the 3DV-ANA as compared to the CNTL. The results demonstrate that the improved initial conditions of the mesoscale model using 3DVAR enhanced the location and amount of rainfall over the Indian monsoon region. Model verification and statistical skill were assessed with the help of available upper-air sounding data. The objective verification further highlighted the efficiency of the data assimilation system. The improvements in the 3DVAR run are uniformly better as compared to the CNTL run for all the three cases. The mesoscale 3DVAR data assimilation system is not operational in the weather forecasting centers in India and a significant finding in this study is that the assimilation of Indian conventional and non-conventional observation datasets into numerical weather forecast models can help improve the simulation accuracy of meso-convective activities over the Indian monsoon region. Results from the control experiments also highlight that weather and regional climate model simulations with coarse analysis have high uncertainty in simulating heavy rain events over the Indian monsoon region and assimilation approaches, such as the 3DVAR can help reduce this uncertainty.     

GNSS反演资料在GRAPES_Meso三维变分中的应用

为了进一步提高GRAPES_Meso的分析和预报效果,该文在GRAPES_Meso三维变分同化系统中建立了同化GNSS/RO反演的大气资料的观测算子,实现了对GNSS/RO反演的大气资料的同化应用,并通过2013年7月1个月的同化和预报试验分析了GNSS/RO反演大气资料对GRAPES_Meso模式系统分析和预报的影响。结果表明:增加了GNSS/RO反演大气资料的同化后,GRAPES_Meso位势高度场的分析误差明显减小,平均分析误差减小约8%,预报误差略有减小,平均预报误差减小约1%;湿度场的分析误差和预报误差变化不明显,常规观测资料稀少的青藏高原地区的降水预报技巧有所提高,小雨到大雨的ETS (equitable threat score) 评分提高约0.01,对全国及其他分区的降水预报技巧总体上有正效果。     

The impact of Global Positioning System data on the prediction of an extratropical cyclone: an observing system simulation experiment

In this paper, we report a series of observing system simulation experiments that we conducted to assess the potential impact of Global Positioning System/meteorology (GPS/MET) refractivity data on short-range numerical weather prediction. We first conducted a control experiment using the Penn State/NCAR mesoscale model MM5 at 90-km resolution on an extratropical cyclone known as the ERICA (Experiment on Rapidly Intensifying Cyclones over the Atlantic) IOP 4 storm. The results from the control experiment were then used to simulate GPS/MET refractivity observations with different spatial resolution and measurement characteristics. The simulated refractivity observations were assimilated into an 180-km model during a 6-h period, which was followed by a 48-h forecast integration. Key findings can be summarized as follows:

• The assimilation of refractivity data at the 180-km resolution can recover important atmospheric structures in temperature and moisture fields both in the upper and lower troposphere, and, through the internal model dynamical processes, also the wind fields. The assimilation of refractivity data led to a considerably more accurate prediction of the cyclone.

• Distributing the refractivity randomly in space and applying a line averaging did not alter the results significantly, while reducing the spatial resolution from 180 km to 360 km produced a moderately degraded result. Even at the 360-km resolution, the GPS-type refractivity data still have a notable positive impact on cyclone prediction.

• Restricting the refractivity data to altitude 3 km and above considerably degraded its impact on cyclone prediction. This degradation was greater than the combined effects of distributing the refractivity data randomly, performing line averaging, and reducing the resolution to 360 km.

These results showed that the GPS/MET refractivity data is likely to have a significant impact on short-range operational numerical weather prediction. The random distribution and line averaging associated with the inherent GPS occultation do not pose a problem for effective assimilation. On the other hand, these results also argue that we need to improve the GPS/MET retrieval algorithm in order to recover useful data in the lower troposphere, and to increase the number of low-earth-orbiting satellites carrying GPS receivers in order to increase the density of GPS soundings, so that the potential impact of GPS/MET refractivity data on numerical weather prediction can be fully realized.     

Determination of moisture changes prior to the onset of south-west monsoon over Kerala using NOAA/TOVS satellite data

Summary The temperature and moisture data from TIROS operational vertical sounder (TOVS) are examined to obtain humidity parameters like, mid and upper tropospheric water vapour, and scale height of water vapour. Their usefulness in characterizing the onset of south-west (SW) monsoon over India is studied. The NOAA satellite data (finished product) with a resolution of 2.5° lat/lon are used to obtain these parameters during and prior to the SW monsoon season over selected regions during 1979 to 1985. The pentad averaged values in the western Indian Ocean showed an increase in scale height of water vapour and mid-tropospheric moisture (700–500 mb) over about 8 to 10 days prior to the onset over Kerala coast. The association of the moisture flux across the Indian Ocean and the rainfall over Kerala coast has also been examined. Results showed that the gradient of middle level moisture is stronger in the case of rainfall deficit years.With 13 Figures     

Observation system experiments for Typhoon Jangmi (200815) observed during T-PARC

In this study, the impact of various types of observations on the track forecast of Tropical Cyclone (TC) Jangmi (200815) is examined by using the Weather Research and Forecasting (WRF) model and the corresponding three-dimensional variational (3DVAR) data assimilation system. TC Jangmi is a recurving typhoon that is observed as part of the THORPEX Pacific Asian Regional Campaign (T-PARC). Conventional observations from the Korea Meteorological Administration (KMA) and targeted dropsonde observations from the Dropwindsonde Observations for Typhoon Surveillance near the Taiwan Region (DOTSTAR) were used for a series of observation system experiments (OSEs). We found that the assimilation of observations in oceanic areas is important to analyze environmental flows (such as the North Pacific high) and to predict the recurvature of TC Jangmi. The assimilation of targeted dropsonde observations (DROP) results in a significant impact on the track forecast. Observations of ocean surface winds (QSCAT) and satellite temperature soundings (SATEM) also contribute positively to the track forecast, especially two- to three-day forecasts. The impact of sensitivity guidance such as real-time singular vectors (SVs) was evaluated in additional experiments.     

Analyses of Orissa Super Cyclone using TRMM (TMI), DMSP (SSM/I) and OceanSat-I (MSMR) Derived Data

The super cyclone in October 1999 was the most intense tropical cyclone in the last century in Orissa, a coastal state in India. This state was battered for more than two days by strong winds and intense rain killing thousands of people. The main objective of this study is to examine the impact of total precipitable water content (TPWC) and surface wind speed data from sensors on board the Tropical Rainfall Measuring Mission (TRMM), Defense Meteorological Satellite Project (DMSP), and Indian Remote Sensing Satellite (OceanSat-I) satellites on the data assimilation system at NCMRWF, New Delhi during the Orissa cyclone period. Comparison of various assimilation experiments suggests that the utilization of TRMM Microwave Imager (TMI) data in the assimilation produced the best analyses. However, in all the forecasts, the storm was predicted to weaken and did not have a reasonably good track. Assimilation experiments with the other two satellite data showed the cyclone track much to the south of the observed track and also it was a weak storm. Biases in the data, when compared with each other, are evident in the analyses also. Better analyses are obtained when the satellite data are used in the originally obtained resolution than when reduced by averaging. A forecast experiment with assimilated data, utilizing the Cloud Motion Vectors (CMVs) from METEOSAT along with TMI data, produced the best forecast among all the experiments. However, the forecast quality was poor. A high-resolution data assimilation experiment was carried out to see the impact of model resolution on the analyses of the cyclone. The strength of the cyclone further increased when higher resolution TMI data were included. The study highlights the need for more satellite data over the Indian Ocean, where conventional data coverage is too poor to define the vertical structure of the atmosphere.     

TAMDAR Observation Assimilation in WRF 3D-Var and Its Impact on Hurricane Ike (2008) Forecast

This study evaluates the impact of atmospheric observations from the Tropospheric Airborne Meteorological Data Reporting (TAMDAR) observing system on numerical weather prediction of hurricane Ike (2008) using three-dimensional data assimilation system for the Weather Research and Forecast (WRF) model (WRF 3D-Var). The TAMDAR data assimilation capability is added to WRF 3D-Var by incorporating the TAMDAR observation operator and corresponding observation processing procedure. Two 6-h cycling data assimilation and forecast experiments are conducted. Track and intensity forecasts are verified against the best track data from the National Hurricane Center. The results show that, on average, assimilating TAMDAR observations has a positive impact on the forecasts of hurricane Ike. The TAMDAR data assimilation reduces the track errors by about 30 km for 72-h forecasts. Improvements in intensity forecasts are also seen after four 6-h data assimilation cycles. Diagnostics show that assimilation of TAMDAR data improves subtropical ridge and steering flow in regions along Ike’s track, resulting in better forecasts.     

The active tropical cyclone season of 2005–2006 over Northwest Australia: Operational model performance and high resolution case studies

Summary There are three main aims of this study. First, the main features of the active 2005–2006 Australian region tropical cyclone (TC) season are summarized, with particular emphasis on the northwest Australian region. Second, an assessment is made of the skill of the available operational global and regional numerical weather prediction (NWP) models for three of the most significant TCs (TCs Clare, Glenda and Hubert), each of which made landfall on the northwest coast of Australia. Third, high-resolution numerical modelling simulations of these same three TCs are described in detail. The numerical weather prediction (NWP) model used here was developed at the University of Oklahoma, and in this study it utilises initial and boundary conditions obtained from archived analyses and forecasts provided by the Australian Bureau of Meteorology, as well as a 4D-Var data assimilation scheme to ingest all available satellite data. The high-resolution numerical model is multiply two-way nested, with the innermost domain having a resolution of 5 km. It was found that unlike the operational models, which were restricted by relatively low resolution and less data, the high resolution model was able to capture most of the major features of all three TC lifecycles including development from initial tropical depressions, intensification, and their tracks, landfall, and associated rainfall and wind fields.     

Organization of convection and monsoon forecasts

Summary In this paper we address the issue of monsoon forecasts in relation to the organization of convection. Given a physical initialization procedure, within a data assimilation, it is possible to use the detailed distribution of rainfall from mesoconvective precipitating elements to define the initial state of a global model. If that is carried out using a very high resolution model then the initial state can carry within it an organization of convection within the resolvable scales. Then the impact of physical initialization on the maintenance and prediction of tropical weather such as the monsoon can be determined. Lacking such an initialization, one can expect the convectively driven energetics to be biased, and a slow degradation of the forecasts can follow. Several examples of forecasts at different resolutions are discussed here. The main findings of this study are that improved forecast results are obtained when physical initialization is invoked where the observed rain and the model resolution are comparable, i.e. the footprint of the highest resolutions rainfall estimates obtained from satellite based data sets (principally we use the SSM/I instrument over the oceans). At this resolution, we note that the model is able to carry an organization of convection in the initialization and in the forecasts through the medium-range time scale.We have compared our results of monsoon studies at a resolution T255 with those at resolution T62. The transform grid separation at the resolution T255 is approximately 50 km and at the resolution T62, it is approximately 200 km. We find that the model at the higher resolution (T255) performs better and has more realistic energy conversions for the convectively driven synoptic scale monsoon.An organization of convection, at the synoptic scales, is not seen in the forecasts at lower resolutions, T62, where the rainfall patterns are generally much broader and tend to be more zonal. Such organization appears more realistic at the resolution T255. Variances of the energy conversion, calculated in the two-dimensional spectral space, from physically initialized short range forecasts at the higher resolution are seen to be largest on the scales of the monsoon. Similar calculations for the reanalyzed fields at lower resolutions show the spectral distribution of variances to be biased towards local Hadley scale overturnings.With 12 Figures     

bbGPS/PWV资料三维变分同化改进MM5降水预报连续试验的评估

利用区域地基GPS网反演的高时空密度的大气垂直方向水汽总量,也称为可降水量（PWV）,可大大弥补常规探空探测水汽资料的不足。为了全面评估区域GPS网PWV资料同化对业务数值天气预报改进程度的目的,在个例研究分析的基础上,进行了连续38天的GPS/PWV资料三维同化（3D-Var）改进数值业务预报的试验。研究方法是根据长江三角洲地区GPS气象网在2002年梅雨和盛夏季节观测的刖资料,通过三维变分同化建立中尺度数值预报模式MM5的初始场,逐日作出长江三角洲地区24小时的降水量预报。以6小时累积雨量为对象,与未同化GPS／刖资料的MM5的相应预报比较,通过多种评分方法,评估了GPS／PWV资料改进MM5降水预报的效果。结果表明GPS／PWV资料同化后的MM5降水预报能力在大部分时间和大部分地区都有所提高,主要是伪击率有较明显的下降,对小范围降水预报的改进更为明显。预报明显改进的区域恰好位于GPS站填补常规探空站间距较大的地区。