Review of the Local Analysis and Prediction System Project：的局部分析和预报系统项目的审查.ppt

《Review of the Local Analysis and Prediction System Project：的局部分析和预报系统项目的审查.ppt》由会员分享，可在线阅读，更多相关《Review of the Local Analysis and Prediction System Project：的局部分析和预报系统项目的审查.ppt（74页珍藏版）》请在三一办公上搜索。

1、LAPS Engineering Overview,LAPS-Local Analysis and Prediction System,John McGinley and Linda Wharton,LAPS Engineering Overview,LAPS-Local Analysis and Prediction System,The goal of LAPS is to ingest spatially and temporally diverse data and output analysis grids.Outline of Engineering Overview:Data I

2、ngest:Sources and ProcessesAnalysis ProcessesSystem/Software ArchitectureLAPS InstallationLAPS LocalizationRunning LAPSUse of LAPS Analysis Grid Files,LAPS Data Ingest Overview,LAPS-Local Analysis and Prediction System,A main goal of the LAPS ingest is to keep ingest code separate from analysis code

3、.This is done by pre-processing ingest data into“intermediate”files that have formats that are expected by the analysis code.The way to incorporate your locally available data into LAPS is to pre-process it into the“intermediate”file format LAPS expects.The intermediate files are in netCDF or ASCII

4、formats.,LAPS Data Ingest,LAPS-Local Analysis and Prediction System,Data Ingest Sources,LAPS-Local Analysis and Prediction System,LAPS can ingest the following data:Gridded Background Models(RUC,NAM,GFS)Satellite(imager,sounder,retrieved soundings,and cloud-top pressure)Surface DataRAOB/Dropsonde/Ra

5、diometerRadarPIREPS&ACARS from aircraftWind Profiler/RASSGPSCloud Drift Winds,Data Ingest Processes-Background,LAPS-Local Analysis and Prediction System,LAPS uses Gridded Model“Background”Data as a model first guess for the LAPS analysis.The Background model must cover the entire LAPS grid both vert

6、ically and horizontally.The Background model may be at a coarser resolution than LAPS.The model data is mapped onto the LAPS grid,both vertically and horizontally.If necessary,the ingest process will temporally interpolate background data so there is data available for each LAPS cycle.,Data Ingest P

7、rocesses-Background,LAPS-Local Analysis and Prediction System,The Data read from the background model are:3DSurfaceheightheighttemperaturetemperaturespecific humidityspecific humidityu,v windsu,v windsomega windsurface pressureMean Sea Level pressure,Data Ingest Processes-Background,LAPS-Local Analy

8、sis and Prediction System,The LAPS process that ingests the Background data is lga.exe The output from lga.exe is stored in netCDF files,with 3D data in the lga output directory and 2D surface data in the lgb output directory.The variables output into the 3D lga file are:height,temperature,specific

9、humidity,u and vwinds,and wind omega.The variables output into the surface lgb file are:temperature,specific humidity,u and v winds,dewpoint,surface pressure,MSLP,and reducedpressure.,Data Ingest Processes-Satellite,LAPS-Local Analysis and Prediction System,lvd_sat_ingest is the process that ingests

10、 GOES satellite data.LAPS makes use of the following satellite channels:visible 3.9 6.7(water vapor)11.2(window IR)12.0 If the raw satellite data does not cover the entire LAPS domain,the missing area is flagged using the LAPS missing_data value.,Data Ingest Processes Surface Data,LAPS-Local Analysi

11、s and Prediction System,LAPS ingests surface data from two netCDF file formats,METAR/SYNOP and LDAD.METAR files are generated hourly by the National Weather Service(NWS)and contain data collected by meteorologists at each Weather Forecast Office(WFO)and NWS automated weather stations.LDAD files cont

12、ain data from diverse sources,either automatically or by non-meteorologists,and have the potential to store more varied data,such as ship and buoy.The LAPS process that ingests surface data is obs_driver.Once the surface data is ingested,gross“climatological”quality control(qc)checks are performed.o

13、bs_driver combines the data from METAR and LDAD files and the output is an ASCII file in the lso output directory,Data Ingest Processes Surface Data,LAPS-Local Analysis and Prediction System,The additional Data read from the LDAD netCDF file are:wet bulb temperaturesolar radiationcloud base heightse

14、a surface temperaturelow level cloud typebuoy/ship-typemid level cloud typebuoy/ship true directionhigh level cloud typebuoy/ship true speedprecipitation ratewave periodprecipitation typewave height precipitation intensitymax temp recording periodtime since last precipmin temp recording period Since

15、 many data sources only report sporadically,the time the data last changed is recorded for:temperature,relative humidity,station pressure,wind speed,wind direction,wind gust,solar radiation,Data Ingest Processes Blacklisting,LAPS-Local Analysis and Prediction System,LAPS has implemented a“blacklist”

16、for surface data stations located in either METAR or LDAD input files.This allows obs_driver to skip stations with known bad variables(one or several)or to skip a station completely.The current data that can be blacklisted are:temperature,dewpoint,relative humidity,wind,altimeter,station pressure,MS

17、L pressure,visibility,clouds(all layers),precipitation,snow cover,solar radiation,soil/water temperature,soil moisture.The Blacklist.dat file is an ASCII file that is installed in the data/static directory.A Blacklist.example file comes with LAPS,and formatting information can be found in the LAPS R

18、EADME file.,Data Ingest Processes Vertical Soundings,LAPS-Local Analysis and Prediction System,Sounding data is used if the observations lie in the time window of the laps cycle time and if the data is available.Sounding data sources are RAOB,dropsonde,satellite sounding and radiometer.Data read in

19、from the RAOB sounding files are:latitude,longitude,elevation,release time,synoptic time,station name,wmoid At Mandatory levels:height,pressure,temperature,dewpoint depression,wind direction,wind speed At Significant Wind levels:height,wind direction,wind speed At Significant Temperature levels:pres

20、sure,dewpoint depression,temperature,Data Ingest Processes-Radar,LAPS-Local Analysis and Prediction System,LAPS makes use of Wideband Radar Data(WSR-88D,Level II),using 3D reflectivity and velocity in polar coordinates.Up to 20 Wideband radars may be processed within the domain.LAPS also uses Narrow

21、band Radar Data(WSR-88D,Level III),using a single level of reflectivity in polar coordinates.Up to 9 Narrowband radars may be processed within the domain.The Data read in from the radar files are:site name,latitude,longitude,elevation,reflectivity,velocity,spectrum width,gate spacing for velocity an

22、d reflectivity,range to first gate for velocity and reflectivity,nyquist velocity,number of radials,radial azimuth and elevation angles,LAPS Analysis Processes,LAPS-Local Analysis and Prediction System,windsurfacetempcloudhumidderivaccumsoil,LAPS Analysis-Legend,LAPS-Local Analysis and Prediction Sy

23、stem,lso,lrs,analysis,lvd,lgb,LC3,LWM,lga,LSX,OUTPUT File=dark blue box,upper case letters,previous OUTPUT File=dark green box,upper case letters,Required input,Optional input,Intermediate file=light blue box,lower case letters,Analysis Flow Chart(from FMI),LAPS-Local Analysis and Prediction System,

24、Wind Analysis,LAPS-Local Analysis and Prediction System,The Wind Analysis is generated using surface observations,profiler data,cloud drift winds and aircraft reports.Background model grids are used as a first guess and to do quality control on new observations.Time tendencies from the background mo

25、del are applied to the aircraft/cloud-drift wind reports when they are taken before or after the nominal analysis time.,Wind Analysis,LAPS-Local Analysis and Prediction System,Quality control within the Wind Analysis rejects any observations deviating from the background by more than a threshold dep

26、ending on observation type:ACARS 10 m/sCloud-Drift winds 10 m/sProfiler 22 m/sDoppler Radar 12 m/sOther 30 m/s,Wind Analysis,LAPS-Local Analysis and Prediction System,The wind analysis is done in three steps.The first step analyzes the non-radar data with the background wind field using a multiple i

27、teration successive correction technique.For the second step,the first step results are used as the background.The data used includes non-radar data;any grid-points with multiple-Doppler radial velocities are also mixed in.Radial velocities are taken from the Doppler radars after dealiasing and othe

28、r quality control steps are done.If two or more radars illuminate a given grid-point,a full wind-vector is constructed from a combination of the radial velocities and the preliminary non-radar analysis.This is done via a successive insertion process,beginning with the background(non-radar analysis),

29、then followed with the radial velocity from each radar in sequence.,Wind Analysis,LAPS-Local Analysis and Prediction System,For the final step the background field comes from the result of the second step.All point data is now used,including grid-points illuminated by only a single radar.The tangent

30、ial component for each radar observation is estimated by using the background from the previous step(i.e.non-radar data and/or multi-radar data).The omega field is calculated by kinematically integrating the horizontal wind divergence.The lower boundary condition is specified by the surface wind and

31、 terrain gradient.The LAPS process that runs the Wind Analysis is wind.exe.,Wind Analysis,LAPS-Local Analysis and Prediction System,Must have to run data:surface data(lso)background(lga)Optional data:vertical sounding(snd)profiler(pro)aircraft(pin)cloud drift winds(cdw)3D temperature(LT1)surface ana

32、lysis(LSX)doppler winds(Vxx),Wind,snd,analysis,pin,pro,LT1,LSX,lga,cdw,lso,LW3 LWM,Wind Analysis Output,LAPS-Local Analysis and Prediction System,The output from wind is stored in netCDF files in the lw3 and lwm output directories.LW3 Output variables are:3D wind u-component 3D wind v-component 3D o

33、mega LWM Output variables are:surface wind u-component surface wind v-component,Surface Analysis,LAPS-Local Analysis and Prediction System,The surface package collects surface data from the lso intermediate data file,IR brightness temperatures from the lvd file,and fields from selected background mo

34、dels.Data is placed on the LAPS grid and a three quality control checks are performed on the observations.This first checks the observations against climatologically reasonable ranges.Second,the observations(most fields except wind)are checked to see which ones are outliers(at 5 standard deviations)

35、relative to the average observation value in the domain.As a further check,the temperatures,dewpoints and MSL pressures are checked to see if they deviate from the background field by more than a threshold absolute amount.,Surface Analysis,LAPS-Local Analysis and Prediction System,The third check fo

36、r all analyzed fields(except visibility)is within the spline routine,that rejects stations deviating from the background by more than a threshold number of standard deviations of the observation increments.This threshold can be independently adjusted(i.e.tightened or loosened)for each field via the

37、surface_analysis.nl namelist.The next step in the analyses is done with a successive correction technique similar to the 3-D wind and temperature analyses.,Surface Analysis,LAPS-Local Analysis and Prediction System,Observation increments are used for T,Td,U,V,MSL,P and straight observations are used

38、 for visibility.The temperature and dewpoint observations are also corrected for deviations of the station elevation from the LAPS terrain.Standard lapse rates are applied to this elevation difference.The analysis innovation is constrained to vary from the background by no more than the magnitude of

39、 the observation rejection threshold discussed above.This helps prevent overshooting(ballooning)of gradients into data sparse areas.A land fraction term is factored into the weighting whenever the observation and grid point are on either sides of a 0.01 land fraction threshold.This helps prevent sit

40、uations such as heating over the land having undue effects over the water areas.This weight is applied mainly to the T,Td,U,and V fields.,Surface Analysis,LAPS-Local Analysis and Prediction System,For pressure analysis,three fields are computed including reduced pressure(P)at reference height redp_l

41、vl,surface pressure(PS),and mean sea level pressure(MSL).Background pressure fields come from the lgb files.The MSL background is used as read in upon input.The(PS)background is converted from the background model terrain to the LAPS terrain within the lgb file.The(P)background is generated by reduc

42、ing the(PS)background to the reference analysis height redp_lvl using Poissons equation.,Surface Analysis,LAPS-Local Analysis and Prediction System,Continuing the pressure analysis,the altimeter setting observations are converted to station pressures using the standard atmosphere.Station pressure ob

43、servations are in turn converted to reduced pressure using Poissons equation.The(P)analysis uses the(P)background plus the reduced pressure observation increments.The(P)analysis then uses variational techniques to constrain the surface winds and reduced pressures(P)to the full equations on motion.In

44、 contrast,mean sea level pressure(MSL)is a direct analysis of the MSLP observation increments together with the model background MSL field.The station pressure analysis(PS)is calculated using the model background gridded PS field,together with the deviations of the MSLP analysis from the MSLP backgr

45、ound.,Surface Analysis,LAPS-Local Analysis and Prediction System,Visibility is arrived at by first analyzing the surface visibility observations.A second step is applied to decrease the visibility in areas that have high RH and are near the cloud base that is given by the cloud analysis in the previ

46、ous time cycle.Several derived variables are calculated before the output file is written.Also,a dependent data validation is done by interpolating several variables back to the observation locations and comparing the analysis to the observations.The LAPS process that runs the Surface Analysis is la

47、ps_sfc.x.,Surface Analysis,LAPS-Local Analysis and Prediction System,User can define if must have to run(default=no):surface data(lso)Optional data:RASS(lrs)satellite(lvd)background(lga,lgb)cloud(LC3)previous mean wind(LWM)previous surface analysis(LSX),Surface,lso,lrs,analysis,lvd,lgb,LC3,LWM,lga,L

48、SX,Surface Analysis Output,LAPS-Local Analysis and Prediction System,The output from laps_sfc is stored in netCDF in the lsx output directory.LSX Output variables are:surface u and v winds vertical velocity vorticity dewpoint temperature temperature relative humidity mean sea level pressure surface

49、pressure reduced pressure potential temperature mixing ratio surface wind speed temperature advection moisture advection ground temperature moisture convergence divergence visibility equivalent potential temperature fire danger potential temperature advection heat index,Temperature Analysis,LAPS-Loc

50、al Analysis and Prediction System,The Temperature Analysis analyzes data from the model background,vertical soundings,satellite soundings,surface temperature analysis and RASS data.Quality control is applied to the temperature soundings.If any level in a sounding differs from the model background by