import os.path import os from nansat import Nansat, Domain from openwind import SARWind import numpy as np from datetime import datetime import dateutil.parser NCEP_wind_dir = ['/Data/sat/auxdata/model/ncep/gfs/',] class AsarImage(Nansat): ''' Read parameters from RADARSAT2 file ''' def __init__(self, fileName, logLevel=10): ''' Read attributes from file name RS2_20110430_145008_0008_F1_HHHV_SGF_130933_4554_5368385 Set values to the dictionary data ''' #init nansat object (open GDAL dataset, get mapper, etc) add new #functions Nansat.__init__(self, str(fileName), mapperName='asar', logLevel=logLevel) #set values to the attribute data where it is availabe fro ImagesCatalog self.set_metadata('name', "'%s'" % self.name) self.set_metadata('path', "'%s'" % self.path) self.set_metadata('sensor', "'asar'") self.set_metadata('sensstart', "'%s'" % self.get_time()[0].strftime('%Y-%m-%d %H:%M:%S.%f')) self.set_metadata('border', self.get_border_postgis()) def process_web(self, opts=None): ''' NRT Processing of Radarsat includes: Visualisation of small quick look Visualisation of large quick look Generating map Input: ------ opts: dictionary with processing options ''' oBaseFileName = self.get_metadata('name').strip('"').strip("'") qlName = opts['mapDir'] + '/' + oBaseFileName + '_.jpg' mapName = opts['mapDir'] + oBaseFileName + '_.png' """ # 1. Generate small quicklook if not os.path.exists(qlName): self.resize(width=300) try: self.write_figure(qlName, clim='hist', ratio=0.9, cmapName='gray') except: self.logger.error('Cannot write preview of %s' % self.fileName) self.resize() """ # 3. Generate map if not os.path.exists(mapName): self.write_map(mapName) return 0 def process_wind(self, opts=None): ''' Compute wind speed and direction using openwind Input: ------ opts: dictionary with processing options ''' oBaseFileName = self.get_metadata('name').strip('"').strip("'") oBasePath = self.get_metadata('path').strip('"').strip("'") mapName = opts['mapDir'] + oBaseFileName + '_.png' resultName = opts['mapDir'] + oBaseFileName + '_.tif' if not os.path.exists(mapName): sar2wind = self.get_wind() if sar2wind is None: return 1 else: # if wind direction and speed were calculated, create the map if sar2wind.has_band('winddirection') and sar2wind.has_band('windspeed'): title = ('Envisat ASAR Wind Speed ' + sar2wind.get_metadata()['MPH_SENSING_START'].split('.')[0]) # Save the map sar2wind.save_wind_map_image(mapName, landmask=True, drawgrid=False, cbar_fontsize=8, vmin=0, vmax=30, tight=True, title=title, edgecolor=0.0390625, quiver_width=0.002, quiver_X=0.8, quiver_Y=0.1, quiver_U=10, quiver_label='10 m/sec', quiver_fontproperties={'size': 8}, quiverScaleCriteria = {'%s<=10.0': 200, '10.0<%s<=20.0': 300, '20.0<%s': 400} ) ''' # Create Nansat object with wind speed and direction n = Nansat(domain=sar2wind) # add wind speed band n.add_band(sar2wind['windspeed'], parameters = sar2wind.get_metadata(bandID='windspeed')) # change wind direction "from" to "to" winddirection = np.mod(sar2wind['winddirection'] + 180, 360) winddirParms = sar2wind.get_metadata(bandID='winddirection') winddirParms['long_name'] = 'Wind direction (to)' winddirParms['standard_name'] = 'wind_to_direction' winddirParms['wkv'] = 'wind_to_direction' # add wind direction band n.add_band(winddirection, parameters = winddirParms) # export to GeoTiff n.export(resultName, driver='GTiff') ''' else: return 1 return 0 def process_wind_thredds(self, opts=None): sar2wind = self.get_wind() if sar2wind is None: return 1 else: # get only wind data sar2wind.vrt.tps = False self.get_wind_only(sar2wind) # create file name dateString = {} for iKey in ['sar_start_date', 'sar_stop_date']: dt = datetime.strptime(sar2wind.get_metadata(iKey), '%d-%b-%Y %H:%M:%S.%f') dateString[iKey] = str('%4d%02d%02dT%02d%02d%02d' %(dt.year, dt.month, dt.day, dt.hour, dt.minute, dt.second)) oFileName = opts['threddsDir']% (dateString['sar_start_date'], dateString['sar_stop_date']) # if the file does not exist yet, process ... if not os.path.exists(oFileName): # reproject sar2wind onto given domain domain = Domain(srs=opts['domain']['srs'], ext=opts['domain']['ext']) sar2wind.reproject(domain, use_geolocationArray=False) createdTime = None if len(opts['additionalMetadata']) != 0: metadata, createdTime = self.set_additionalMetadata(sar2wind, opts['metadata'], opts['additionalMetadata']) try: time = filter(None, self.get_sar_time(sar2wind, timeName=opts['timeName'])) sar2wind.export2thredds(oFileName, maskName = opts['maskName'], rmMetadata=opts['rmMetadata'], bands=opts['bands'], metadata=metadata, time=time, createdTime=createdTime) except: return 1 return 0 def get_wind(self): sar = Nansat(self.fileName) sar.vrt.tps = True sar.crop(lonlim=[10, 76], latlim=[63, 85]) ncepFileName = self.get_ncep(self.fileName) try: ncep = Nansat(ncepFileName) except: return None else: # Set vrt.tps for accuracy of reprojection ncep.crop(lonlim=[10, 76], latlim=[63, 85]) ncep.vrt.tps = True try: sar2wind = SARWind(sar, wind_direction=ncep) except: return None else: if (sar2wind.has_band('U') and sar2wind.has_band('V')): # add mask self.add_mask(sar2wind) # reproject GCPs gcps = sar2wind.vrt.dataset.GetGCPs() clon = 0 clat = 0 k = 0 for gcp in gcps: clon += gcp.GCPX clat += gcp.GCPY k += 1 clon /= k clat /= k srs = '+proj=stere +datum=WGS84 +ellps=WGS84 +lon_0=%f +lat_0=%f +no_defs' % (clon, clat) sar2wind.vrt.reproject_GCPs(srs) return sar2wind else: return None def get_ncep(self, asarFileName): # Search for a ncep file in NCEP_wind_dir corresponding to the given sar data NCEP_wind_dir = ['/Data/sat/auxdata/model/ncep/gfs/',] n = Nansat(asarFileName) sigma0_bandNo = n._get_band_number({'standard_name': 'surface_backwards_scattering_coefficient_of_radar_wave', 'polarization': 'VV'}) SAR_image_time = n.get_time(sigma0_bandNo).replace(tzinfo=None) estr = 'Make sure one of these directories are connected:\n\n' for dir in NCEP_wind_dir: estr = estr + dir + '\n' if os.path.exists(dir): ndir = dir break if not 'ndir' in locals(): raise Exception, estr basehour = np.floor((SAR_image_time.hour + SAR_image_time.minute / 60.0 + 3.0 / 2.0)/6.0)*6.0 basehour = np.min([18, basehour]) if SAR_image_time.hour + SAR_image_time.minute / 60.0 - basehour > 1.5: forecasthour = 3 else: forecasthour = 0 nfile = os.path.join(ndir, 'gfs%4d%02d%02d/' %(SAR_image_time.year, SAR_image_time.month, SAR_image_time.day), 'gfs.t%.2dz.master.grbf%.2d' %(basehour, forecasthour) ) if os.path.exists(nfile): print 'Found ' + nfile + ' file in NERSC' ncepFileName = nfile else: print 'Could not find ' + nfile +' file in NERSC' ncepFileName = None return ncepFileName def add_mask(self, data): mask = 64 * np.ones(data.shape()).astype('int8') # land mask watermask = data.watermask(tps=True) mask[watermask[1]==2] = 2 # no data mask mask[data[1]==0.0] = 0 data.add_band(array=mask, parameters={'wkv':'land_mask', 'name': 'mask', 'long_name': 'L2-mask', 'standard_name': 'mask'}) def get_wind_only(self, data): # remove bands except for U, V and mask delBandList = list(np.arange( 1, data.vrt.dataset.RasterCount + 1)) delBandList.remove(data._get_band_number('U')) delBandList.remove(data._get_band_number('V')) delBandList.remove(data._get_band_number('mask')) data.vrt.delete_bands(delBandList) # set start and stop date-time to global and band metadata dtMetadata = {'sar_start_date': data.get_metadata('SPH_FIRST_LINE_TIME'), 'sar_stop_date': data.get_metadata('SPH_LAST_LINE_TIME'), 'FillValue': '-10000' } data.set_metadata(dtMetadata) data.set_metadata(dtMetadata, bandID='U') data.set_metadata(dtMetadata, bandID='V') def get_sar_time(self, data, bandID=None, timeName='time'): time = [] for i in range(data.vrt.dataset.RasterCount): band = data.get_GDALRasterBand(i + 1) try: time.append(dateutil.parser.parse( band.GetMetadataItem(timeName))) except: data.logger.debug('Band ' + str(i + 1) + ' has no time') time.append(None) if bandID is not None: bandNumber = data._get_band_number(bandID) return time[bandNumber - 1] else: return time def set_additionalMetadata(self, data, metadata, additionalMetadata): # get corners of data lonCrn,latCrn = data.get_corners() # get current time (creation time) time = datetime.utcnow().strftime('%Y-%m-%d %H:%M:%S UTC') for iMeta in additionalMetadata: if (iMeta == 'sar_start_date' or iMeta == 'time_coverage_start' or iMeta == 'sar_stop_date' or iMeta == 'time_coverage_end'): if iMeta in 'start': keyName = 'sar_start_date' else: keyName = 'sar_stop_date' dt = datetime.strptime(data.get_metadata(keyName), '%d-%b-%Y %H:%M:%S.%f') metadata[iMeta] =('%4d-%02d-%02d %02d:%02d:%02d UTC' %(dt.year, dt.month, dt.day, dt.hour, dt.minute, dt.second)) elif iMeta == 'geospatial_lat_min': metadata[iMeta] = np.float(min(latCrn)) elif iMeta == 'geospatial_lat_max': metadata[iMeta] = np.float(max(latCrn)) elif iMeta == 'geospatial_lon_min': metadata[iMeta] = np.float(min(lonCrn)) elif iMeta == 'geospatial_lon_max': metadata[iMeta] = np.float(max(lonCrn)) elif iMeta == 'date_created' or iMeta == 'date_modified': metadata[iMeta] = time return metadata, time