int.j.remote sensing,2001,vol.22,no.7,1285±1303
Validation of coastal sea and lake surface temperature measurements derived from NOAA/AVHRR data
X.LI2,W.PICHEL ,P.CLEMENTE-COLO N ,
V.KRASNOPOLSKY§and J.SAPPER
Research and Data Systems Corporation,Room102,E/RA3,WWBG,NOAA
Science Center,5200Auth Road,Camp Springs,Maryland20746-4304,USA;
e-mail:xiaofeng.li@https://www.doczj.com/doc/257038779.html,
NOAA/NESDIS,Room102,E/RA3,WWBG,5200Auth Road,Camp
Springs,Maryland20746-4304,USA
§General Sciences Corporation,6100Chevy Chase Drive,Laurel,Maryland
20707,USA
(Received11January1999;in nal form20December1999)
Abstract.An interactive validation monitoring system is being used at the
NOAA/NESDIS to validate the sea surface temperature(SST)derived from the
NOAA-12and NOAA-14polar orbiting satellite AVHRR sensors for the NOAA
CoastWatch program.In1997,we validated the SST in coastal regions of the
Gulf of Mexico,Southeast US and Northeast US and the lake surface temper-
atures in the Great Lakes every other month.The in situ temperatures measured
by24NOAA moored buoys were used as ground data.The non-linear SST
(NLSST)algorithm was used for all AVHRR SST estimations except during the
day in the Great Lakes where the linear multichannel SST(MCSST)algorithm
was used.The buoy±satellite matchups were made within one image pixel in space
(1.1km at nadir)and1h in time.
For the NOAA-12satellite,the validation results for the three coastal regions (Gulf of Mexico,Southeast US and Northeast US)showed that the mean temper-
ature di V erence between satellite and buoy surface temperature(bias)was about
0.4C during the day and0.2C at night.The standard deviation was about1.0C.
Great Lakes validation results showed a bias less than0.3C during the day.
However,due to the early morning fog situation in the summer months in the
Great Lakes region,the NLSST night algorithm yielded a fairly large bias of
about1.5C.
The same statistics were computed for the NOAA-14satellite measurements.
For the coastal regions,the bias was less than0.2C with a standard deviation
about1.0C.For the Great Lakes region,the bias was about0.4C for both day
and night with a standard deviation about1.0C.
Our study also showed that the NLSST algorithm provides the same order of SST accuracy over all study regions and under a wide range of environmental
conditions.
1.Introduction
The derivation of sea surface temperature(SST)from satellite measurements has been a focus of numerous studies since the early1970s(Anding and Kauth1970, McMillin1975,McMillin et al.1975,Barton1983,Llewellyn-Jones et al.1984,
International Journal of Remote Sensing
ISSN0143-1161print/ISSN1366-5901online2001Taylor&Francis Ltd
1286
X.L i et al.
McMillin and Crosby1984,McClain et al.1985,Walton1988,Barton et al.1989, Minnett1990,Emery et al.1994,Walton et al.1998).The Advanced Very High Resolution Radiometer(AVHRR/2)onboard the NOAA series of Polar-orbiting Operational Environmental Satellites(POES)is primarily designed for SST retrieval and cloud detection.POES satellites known as Advanced Television Infrared Observation Satellites(TIROS-N or ATN)operate as a pair to ensure that the data, for any region of the earth,are no more than6h old.AVHRR has ve channels, two visible channels(channels1and2at0.6and0.9m m,respectively),one short-wavelength infrared channel(channel3at3.7m m),and two long-wavelength infrared channels,the split window channels(channels4and5at11and12m m,respectively). The wavelengths of the three infrared channels are selected in a range of the electro-magnetic spectrum in which the radiation from the earth’s surface and clouds is only weakly attenuated.To determine the actual SST from the AVHRR radiation measure-ments,one must correct for absorption and reemission of radiation by atmospheric gases,predominately water vapour.The split window method,which uses the channel 4and5brightness temperatures to calculate SST,is widely used for atmospheric correction.A summary and comparison of di V erent split window algorithms are given in Barton(1995).
NOAA’s National Environmental Satellite,Data,and Information Service (NESDIS)produces two main types of SST products;i.e.global SST and CoastWatch SST.The global SST suite of products are generated from AVHRR Global Area Coverage(GAC)4km data recorded on-board the POES satellites and downlinked to NESDIS acquisition stations at Wallops Station,Virginia and Fairbanks,Alaska. Global SST measurements are produced at8km resolution with variable spacing from8to25km in cloud-free areas twice per day from each of the two operational POES satellites.The global satellite SST measurements are validated by comparing them with drifting buoy(and TOGA moored buoys in the tropical Paci?c)SST measurements matching within4h and25km.These global satellite SST measure-ments are used to produce SST analyses at grid resolutions from14to100km. CoastWatch SST products are generated from a di V erent data stream,the AVHRR High Resolution Picture Transmission(HRPT)data,broadcast continuously by the POES satellites.The HRPT data have a resolution of1.1km at nadir and are mapped to almost full resolution in the production of CoastWatch AVHRR visible, infrared and SST images.The CoastWatch products are validated by comparison with NOAA moored buoy SST reports using techniques described herein.Figure1 shows the time lines of di V erent operational algorithms used at NOAA/NESDIS. Based on the split window theory,the multichannel SST(MCSST)algorithm was developed and used operationally at NOAA/NESDIS in the early1980s.This algo-rithm assumes that there is a linear relationship between the di V erence of the actual SST and a satellite measurement in one channel and the di V erence of satellite measurements in the split window channels(channel4and5).Therefore,the actual SST can be estimated using brightness temperatures measured with channels4and 5.Walton(1988)considered a non-linear term in the further development of MCSST and developed the cross-product SST(CPSST)algorithm.A simple version of the CPSST algorithm,called the non-linear SST(NLSST)algorithm,was implemented at NESDIS for operational use in April1991.The coe cients for these algorithms are routinely obtained by performing a regression between satellite retrievals and buoy data soon after each satellite’s launch.
AVHRR/SST accuracy for NOAA-12and-14satellites1287
Figure1.Time lines of NOAA series of polar orbiting satellites used for SST and the operational sea surface temperature algorithms used at NOAA/NESDIS.(a)NOAA Global Operation,(b)NOAA CoastWatch Operation.GOSSTCOMP:Global Operational Sea Surface Temperature Computation.MCSST:Multichannel Sea Surface Temperature,the MCSST product started on17November1981.CPSST: Cross-product sea surface temperature,beginning2March1990.NLSST:Non-linear Sea Surface Temperature,NLSST product,starting on10April1991in the global operation and on3June1992in the CoastWatch operation.
oceanic problems.For some applications,relatively low absolute SST accuracy is required as long as high relative accuracy is achieved,i.e.for front and edge detection (Cayula and Cornillon1992,Kahru et al.1995),and in feature tracking and motion detection(Emery and Fowler1991,Breaker et al.1994).However,in some other studies,i.e.climate studies(Harries et al.1983,Yates et al.1985,Cornillon1989),a more stringent absolute SST accuracy,normally less than0.3C,is required.To understand the satellite-derived SST accuracy,scientists have performed various validation e V orts by comparing the AVHRR measurements with moored buoy, drifting buoy and ship measurements in the global ocean as well as in di V erent coastal regions.
For the global GAC SST validation,Strong and McClain(1984)used the data between November1981and February1982and found that the root mean square (rms.)error of the temperature di V erence between satellite and in situ measurement was between0.6and1.8C.Pichel(1991)used3months of a NOAA-11satellite and buoy matchup dataset between March and May1990to validate the NLSST algo-rithm,and found the accuracy had been improved.The global mean satellite±buoy di V erence(or bias)was less than0.3C with a standard deviation of about0.7C. Walton et al.(1998)analysed a9-year time series of satellite±buoy matchups between 1989and1997.They showed that the bias has stayed between0.2and0.4C over
1288
X.L i et al.
SSTs improved from0.8to0.5C for the daytime algorithm but remained about 0.5C for the night-time algorithm.In their study,satellite±buoy matches were constrained to25km and4h.The largest di V erences resulted from the volcanic aerosols from the Mt Pinatubo eruptions in October1992,with a positive bias in the night-time SST measurements observed from the month of the eruptions until June1993.
The matchups made within4h in the global SST validation can be less accurate when a diurnal warming e V ect is considered(Cornillion and Stramma1985,Bo hm et al.1991,Hawkins et al.1993).For regional validations,one needs to set up satellite±buoy matchup datasets at higher spatial and temporal resolutions than are used in global validation studies.So far,there have been only a few studies concerning regional AVHRR SST validation.Pearce et al.(1989)validated the NOAA-7and NOAA-9satellite-derived SST using in situ boat measurements as ground data in the coastal waters o V Western Australia.They compared seven published split window algorithm derived SSTs and found that all algorithms yielded reasonably good results.The rms.error between SSTs calculated with two of the algorithms and their corresponding ship measurements was about0.6C.The bias was between 0.1and0.2C.Robinson and Ward(1989)compared NOAA-7SSTs calculated with the Llewellyn-Jones et al.(1984)split window algorithm with cruise data in the north-east Atlantic Ocean.The ship and satellite measurement agreement was within 1C.Yokoyama and Tanba(1991)compared14published split window algorithms using a matchup dataset in Mutsu Bay in northern Japan for the NOAA-9satellite. They showed that the regional split window algorithm had rms.errors in the range of0.55±0.75C.In their more recent paper,Yokoyama et al.(1993)found that larger satellite retrieval errors appeared to occur when the air±sea temperature di V erence was large.May and Holyer(1993)noticed the satellite SST retrieval error can be as large as1C when the air sea temperature di V erence changes10±12C from the mean conditions in their global dataset.Topliss(1995)reviewed split window algorithms for the NOAA-7,NOAA-9and NOAA-11satellites and developed new regional split window algorithms for the Canadian coastal region.All the above regional SST algorithms are linear SST algorithms.
NOAA/NESDIS uses NLSST rather than regional algorithms for the measure-ment of SST.This avoids the problems of possible discontinuities at the regional boundaries as well as any need for seasonal adjustments within regions(Walton et al.1998).In this study,we use a long-term validation system developed for the NOAA CoastWatch program to validate the accuracy of AVHRR SSTs in the Northeast,Southeast,and Gulf of Mexico coastal regions and lake surface temper-atures in the Great Lakes area for NOAA-12and NOAA-14in1997.In§2, CoastWatch AVHRR data preparation is presented,followed by a description of the validation procedure in§3.In§4,we present validation results.Analysis and discus-sion are in§5,and the conclusions are in§6.
2.NOAA CoastWatch AVHRR data preparation
2.1.Satellite mapped data for CoastWatch
CoastWatch is a NOAA program managed by NESDIS with CoastWatch Nodes located at NOAA laboratories and o ces in eight coastal states.The goal of CoastWatch is to provide satellite and other environmental data and products for near real-time monitoring of US coastal waters in support of environmental science,
AVHRR/SST accuracy for NOAA-12and-14satellites1289 receive them from NESDIS and make them available to a diverse and growing user community of Federal and state environmental resource managers,research scientists, educators,shermen,and marine enthusiasts.Products include polar and geostation-ary satellite infrared,visible,and SST images,as well as ocean colour and Synthetic Aperture Radar(SAR)imagery.Started in1990,with all eight Nodes operating by 1993,CoastWatch had over2100registered users in1997.
Input data for the production of CoastWatch imagery are HRPT1b datasets. These consist of AVHRR detector output from the ve channels of the AVHRR with appended calibration and earth location information.For US east coast,Great Lakes and Gulf of Mexico regions,datasets are received from every satellite pass over the Wallops Station,Virginia reception mask.
During1997,the two operational polar orbiting satellites were NOAA-12and NOAA-14.NOAA-12was launched on14May1991,into a sun-synchronous polar orbit with equator crossing times early in the morning at07:09am descending and in the evening at19:09pm ascending.NOAA-14was launched on30December 1994,into a similar orbit with equator crossing times ascending in the afternoon at 13:43pm local time and descending at night at01:https://www.doczj.com/doc/257038779.html,ually,each CoastWatch region receives satellite coverage four times per day.The local satellite overpass times for NOAA-12and NOAA-14are given in table1.Satellite data from Wallops are transmitted to the NESDIS Central Environmental Satellite Computer System (CEMSCS)in Suitland,Maryland as soon as each satellite overpass is completed. Processing into1b data proceeds automatically as soon as the complete pass has arrived,followed by CoastWatch mapping over each region covered by the satellite pass.
2.2.CoastWatch mapping
The AVHRR NOAA level1b data are mapped to Mercator projection`region’maps covering entire CoastWatch regions.All ve channels,as well as the satellite and solar zenith angles,are mapped at1.1km resolution at nadir.The zenith angle is the angle at a point on the earth between the local normal at that point and a line connecting the point on the earth and the satellite or the sun.The satellite zenith angle is computed using the relation:
sin(h)=(1+H/R)sin(a)(1) where h is the satellite zenith angle,H is the height of the satellite,R is the radius of the earth and a is the scan angle.The scan angle,which is also called the nadir angle,is de?ned as the angle between the line connecting the satellite with the subsatellite point and a line connecting the satellite to a viewed spot on the earth Table1.NOAA-12and NOAA-14local(US east coast)overpass times. Satellite Time GMT Local time(EST) NOAA-12Day21±23Z4:00±6:00pm
Night11±12Z6:00±7:00am NOAA-14Day18±19Z1:00±2:00pm
Night06±07Z1:00±2:00am
X.L i et al.
1290
scan.For AVHRR scans,the scan angle ranges from0to55.4.The scan angle,a, is computed using the relation:
a=(55.4/N)|M N|(2) where M is any given spot number and N is the spot number of nadir.
For NOAA POES satellites,the range of satellite zenith angles can be shown using equation(1)to be from0to68.4.The factor H/R in equation(1)is hardcoded in the image processing programs as0.13,since the NOAA satellite height is about 825km.If there is a signi?cant variation in satellite height,the satellite zenith angles generated are expected to be o V at high zenith angles by1%for every50km di V erence in altitude.At larger satellite zenith angles,the larger atmospheric path length leads to greater attenuation of surface infrared emissions and thus the need for greater correction of AVHRR channel temperatures when calculating SST.Also, since the eld of view increases with satellite zenith angle,there is a greater chance of cloud contamination as zenith angle increases.These e V ects should lead to a decrease in accuracy of SST measurement at high satellite zenith angles.To maintain high accuracy,no SST measurements are attempted at satellite zenith angles above 53.The exception to this rule is in the Gulf of Mexico CoastWatch region where spatial coverage was determined to be more important than absolute accuracy.
Each satellite pixel is calibrated to albedo or equivalent blackbody temperature (correcting for non-linearity in the calibration of channel4and5,see Planet1998) and transformed to a map pixel.Any map pixels left un?lled after all satellite data have been mapped are lled with an average of all the pixels in a55array about the un?lled pixel.To retain the full radiometric precision of the AVHRR instrument, 11bits are used to store the calibrated satellite values(Pichel et al.1991).
2.3.Operational nonlinear SST(NL SST)and multichannel SST(MCSST) algorithms
Once the data have been mapped,then the multiple channels and angles are combined with multichannel algorithms to produce SST and cloud mask imagery. SST imagery is generated with the non-linear NLSST split window algorithm in the US coastal regions.This algorithm utilizes the di V erence between the11and12m m infrared channels to correct for the e V ects of water vapour.Since infrared radiation is absorbed by atmospheric moisture more within the12m m channel than within the11m m channel,the temperature di V erence between these channels is proportional to the amount of water vapour in the atmosphere.The equations also contain a correction for atmospheric path length variation with satellite zenith angle.The linear MCSST split window equation is used to obtain an estimate of the surface temperature for the non-linear term of the NLSST equation.Separate equations are used for day and night data and the equations are satellite dependent.These equations are generated after satellite launch by matching a month’s worth of satellite data with global drifting buoy observations.All matches within25km and4h are used in a regression analysis in order to derive the equations.Because of the global nature of the matchup dataset,the regression equations are usually independent of season, geographic location,or atmospheric moisture content.However,adjustments to the equations have been necessary when instrument or spacecraft environmental changes
AVHRR/SST accuracy for NOAA-12and-14satellites1291 regions of the Earth.The NLSST and MCSST equations used in CoastWatch are given below:
NLSST=A
1(T
11
)+A
2
(T
11
T
12
)(MCSST)+A
3
(T
11
T
12
)(sec h1)A
4
(3)
MCSST=B
1(T
11
)+B
2
(T
11
T
12
)+B
3
(T
11
T
12
)(sec h1)B
4
(4)
where T
11and T
12
are the AVHRR11and12m m channel temperatures in Kelvin;
sec h is the secant of the satellite zenith angle h;NLSST and MCSST are the non-linear and linear multichannel SST retrieval algorithms,respectively,in Centigrade;
A
1A
4
and B
1
B
4
are constant coe cients.A
1
A
4
and B
1
B
4
coe cients for
the NOAA-12and NOAA-14day and night algorithms are given in table2.
Recently,Walton et al.(1998)showed a9-year time series of NOAA-14satellite±buoy monthly bias(i.e.mean satellite±buoy SST di V erence)and scatter(i.e.standard deviation of satellite±buoy SST di V erence)between1989and1998.Their results show that the improvement in the scatter from0.8to0.5C is partly due to improved SST algorithms(from MCSST to NLSST),and partly to the improvements in the cloud detection algorithms.Shenoi(1999)accessed the MCSST and NLSST algo-rithms performance for NOAA-9and NOAA-11satellites.Their results showed that the mean and RMSD values of SST residuals estimated by NLSST are better than those estimated by MCSST for both satellites.
The CoastWatch equations di V er from the global SST equations in three respects:
1.The CoastWatch equations use the MCSST value in the non-linear term rather than an a priori SST estimate obtained from an analysis of past satellite SST data. This means that there is somewhat more noise in the CoastWatch observations. Both the global operation and CoastWatch constrain the value of the a priori SST or the MCSST to the range0±28C.
2.In the Great Lakes,the MCSST value is used as the nal SST value during the day;i.e.a linear equation is used as the operational equation rather than a non-linear equation.In earlier accuracy studies,it was found that the MCSST equations consistently gave slightly more accurate SST measurements than did the NLSST algorithm during the day.
3.The NLSST split-window equation is used for CoastWatch at night rather than the triple-window equation(employing all three infrared channels)which is used in the global operation.For NOAA-12,the3.7m m channel is not used for Table2.NOAA-14and NOAA-12NLSST and MCSST algorithm coe cients used in
CoastWatch SST measurements.
NL SST coe cients
A
1A
2
A
3
A
4
NOAA-14day0.9398130.0760660.801458255.165 NOAA-14night0.9331090.0780950.738128253.428 NOAA-12day0.8769920.0831320.349877236.667 NOAA-12night0.8887060.0816460.576136240.229
MCSST coe cients
B
1B
2
B
3
B
4
NOAA-14day 1.017342 2.1395880.779706278.430 NOAA-14night 1.029088 2.2753850.752567282.240 NOAA-12day0.963563 2.5792110.242598263.006 NOAA-12night0.967077 2.3843760.480788263.940
X.L i et al.
1292
CoastWatch because there is a problem in the calibration of that channel during part of each orbit.For consistency,the NLSST split-window equation is also used for the NOAA-14CoastWatch equations.
2.4.CoastWatch image products
Once SSTs are generated by the NLSST and MCSST algorithms and maps are generated for each CoastWatch region,the CoastWatch mapping system generates a series of`sector’images from the region maps.These sector maps are all512512 pixels in size for selected areas within the region.Sectors are produced at full-resolution for the validation areas shown in gure2.Sector maps can be infrared or visible channels,angles,SST or cloud masks.All the sector products as well as the full-resolution region maps are now being archived.In this study,we use the full resolution images to validate the AVHRR SST product.
A cloud-mask image product useful for interpretation of the SST imagery or for automatic multiday composing of cloud-free pixels is also generated.The algorithm employed is the CLouds from AVHRR(CLAVR)algorithm(Stowe et al.1991).With a series of threshold,uniformity,and channel-di V erence or ratio tests,the CLAVR algorithm determines whether each22pixel array in the region map is clear or cloudy.The cloud maps are generated in the same projection as the SST images and used as aids in determining the clear satellite±buoy matches used in the validation procedure.
3.NOAA CoastWatch validation procedure
The CoastWatch validation system is an interactive,menu-driven,image and data processing system.The system was developed using the Interactive Data Language(IDL)computer language and can be run on both VAX and UNIX
Figure2.CoastWatch high resolution AVHRR data remapped areas used in the CoastWatch validation system.Great Lakes:Lake Huron,Erie and Ontario(GE);Lake Michigan and Huron(GM);Lake Superior(GS).Northeast:Chesapeake Bay(EC);Gulf of Maine(EM);Southern New England(ES).Southeast:East Florida(SE);North Carolina(SN).Gulf of Mexico:Louisiana and Mississippi(ML);Texas(MT);West
AVHRR/SST accuracy for NOAA-12and-14satellites1293 platforms.This system is designed to provide long-term validation for the CoastWatch SST,visible and cloud-mask imagery.The hierarchy chart of the current validation system is presented in gure3.
The National Centers for Environmental Predication(NCEP)provides the buoy data used in the matching procedure.These data are placed in the buoy data le four times a day.The buoy data le also gives the current NOAA moored buoy locations,so an analyst can overlay the buoy positions on the AVHRR imagery. AVHRR imagery is in the CoastWatch format and images are archived at the National Oceanographic Data Center(NODC).The main input for this long-term validation system is the Target Match File(TMF).The TMF is generated by extracting1515pixel array targets of the CoastWatch imagery(i.e.mapped full-resolution AVHRR HRPT imagery including all ve channels,cloud masks and SST)centred at NOAA moored buoy positions on the NOAA/NESDIS Central EnvironMental Satellite Computer System(CEMSCS)mainframe computer.The corresponding buoy data are appended to each target.
The long-term validation system enables an analyst to(1)preview AVHRR
1294
X.L i et al.
images(both infrared and visible channels)to see whether an image contains cloud-free SST measurements at any buoy location,(2)overlay coastlines,grids,buoy locations and AVHRR imagery header information on the images,(3)renavigate the imagery by remapping the image to agree with selected ground control points (Krasnopolsky and Breaker1994),(4)display cloud masks,(5)extract clear33 arrays of CoastWatch SST values centred on each of the buoys in each coastal region,(6)create an output SST match le which contains satellite and buoy SSTs, air temperature,wind and wave information,solar and satellite zenith angles and navigation information,and(7)calculate statistics and make graphic output.Further, cloud screening is done by examining SST and,when necessary,visible imagery.If fog or cloud is suspected,a matchup is not made.Table3shows the locations of the NOAA moored buoys used for validation.
4.Validation results
In1997,the validation was performed in the Gulf of Mexico,Southeast US and Northeast US coastal regions as well as in the Great Lakes region every other month.Both NOAA-12and NOAA-14satellite images were validated.The centre value of the33arrays of SST measurements in the SST match le was taken as the satellite SST.The mean and standard deviation of all the di V erences in each region were then calculated and stored in the SST match le.During1997,there were a total of1829matchups in the three coastal regions,and693matchups in the Great Lakes.The Great Lakes matchups were usually not available in the winter
Table3.NOAA moored buoys used in the AVHRR SST validation.There are a total of24
buoys.
Buoy ID Region https://www.doczj.com/doc/257038779.html,t. 42001Gulf of Mexico88.653325.9283 42002Gulf of Mexico93.567525.8917 42003Gulf of Mexico85.914225.9361 42007Gulf of Mexico88.770030.0900 42019Gulf of Mexico94.999427.8967 42020Gulf of Mexico96.505627.0122 41002South East75.240632.2950 41004South East79.099432.5100 41009South East80.184228.5003 41010South East78.501928.8986 41001North East72.589734.6983 44004North East70.689738.4564 44005North East68.943942.8983 44011North East66.583341.0833 41014North East74.833636.5831 44025North East73.166740.2503 45001Great Lakes87.766448.0481 45002Great Lakes86.418345.3006 45003Great Lakes82.768145.3181 45004Great Lakes86.534247.5458 45005Great Lakes82.398341.6767 45006Great Lakes89.866747.3194 45007Great Lakes87.033342.6833 45008Great Lakes82.415844.2833
AVHRR/SST accuracy for NOAA-12and-14satellites1295 when buoy maintenance was performed.In this study,the Great Lakes matchup dataset consists of data from May,July and September1997.
Due to a calibration error which occurred in NOAA-12night-time passes from early May to early July of1997,large SST measurement biases were found for NOAA-12night-time SSTs.The NOAA-12night-time SST matchup dataset over this period of time was eliminated.There were a total of124and80bad data points for the three coastal regions and the Great Lakes region,respectively.That reduced our matchup dataset to1705and613matchup points.
If the centre value of the33AVHRR SST array was two standard deviations above or below the nine points mean value,this matchup was not used in the statistics calculation.A signi?cant di V erence between the centre and the mean value can occur when there is a thermal front in the33array or some of the33array points are cloud contaminated.After we excluded the matchups beyond two standard deviations,the remaining matchup totals were1602for coastal regions and572for the Great Lakes region,respectively.This means that we included about94%of the matchups from the correctly calibrated dataset in our later analysis.
The number of matches,satellite±buoy bias,and standard deviation of the di V er-ence for all coastal regions(Gulf of Mexico,Southeast US and Northeast US)and the Great Lakes region are given in table4.In addition,the linear correlation coe cients(R)between satellite and buoy measurements are given in table4.The scatter plots of satellite vs buoy measurements for NOAA-12and NOAA-14in the Gulf of Mexico,Northeast US,Southeast US and Great Lakes regions are presented in gures4(a)and(b).
5.Discussion
For the three US coastal regions,NOAA-14AVHRR SSTs calculated with the NLSST algorithm had a bias and standard deviation of0.16C and1.03C for daytime,and0.07C and0.84C for night-time.For NOAA-12daytime SST,the NLSST yields SSTs with a bias of0.43C and a standard deviation of1.00C.The Table4.Mean satellite±buoy SST di V erence(bias)and standard deviation for NOAA-12
and NOAA-14satellites in1997.All24moored buoys matched within one pixel
(1.1km at nadir)and an hour of cloud-free satellite data were used in the validation.
For the Great Lakes region,the MCSST algorithm is used for daytime SST retrievals;
all other measurements are made with the NLSST algorithm.R is the correlation coe cient between AVHRR-derived SST data and buoy measured SST data.
(Satellite±buoy)
Number SST bias SD
Satellite Time Algorithm of matches(C)(C)R CoastWatch Northeast,Southeast and Gulf of Mexico regions
NOAA-14day NLSST4410.16 1.030.9911 NOAA-14night NLSST5020.070.840.9938 NOAA-12day NLSST3740.43 1.000.9909 NOAA-12night NLSST2850.20 1.070.9896 CoastWatch Great L akes region
NOAA-14day MCSST2150.38 1.010.9958 NOAA-14night NLSST1570.410.800.9861 NOAA-12day MCSST1220.260.830.9930 NOAA-12night NLSST78 1.52 1.270.9942
1296
X.L i et al.
Figure4.(a)Scatter plots of satellite vs buoy measurements for NOAA-12and NOAA-14 in the Gulf of Mexico(GM),Northeast US(NE)and Southeast US(SE)coastal region for1997.
bias and standard deviation were0.20C and1.07C for the NOAA-12night-time SSTs(table4).
For the Great Lakes validation,the NLSST-derived NOAA-14SST measure-ments had a bias of about0.4C for both the day and night algorithms,somewhat higher than the coastal SST estimates.For NOAA-12,the MCSST algorithm gave good SST measurements for daytime SST.However,the NOAA-12night algorithm had a SST bias as large as1.52C.The NOAA-12satellite local overpass time is
AVHRR/SST accuracy for NOAA-12and-14satellites1297
Figure4.(b)Scatter plots of satellite vs buoy SST measurements for NOAA-12and
NOAA-14in the Great Lakes region.
months(May,July,September).At this time of the day and this time of the year, there is often uniform low level fog over the lake surface.The fog is so thin and its temperature is so close to(but generally higher than)lake surface temperature,that the cloud test did not detect the presence of fog.Therefore,the NOAA-12night-time algorithm gave large SST biases in the Great Lakes region.We tested the performance of the MCSST algorithm on the NOAA-12night-time matchup data. The MCSST derived bias is0.64C with standard deviation of2.0C.Even though, the MCSST algorithm improved the bias from1.52C to0.64C,the standard
1298
X.L i et al.
Figure5.Scatter plots(satellite minus buoy SST measurements)vs wind velocity for the Gulf of Mexico,Northeast US and Southeast US coastal regions for1997.
we can see there are two schools of matchups.One is at the low temperature end (around4C).The other is between10and22C.From the same gure,we can also see that most of the problems happen when the water temperature is lower than or close to4C(this observation is valid for both NLSST and MCSST).After we restrict the satellite and buoy matchups to those above4C,the bias and standard deviation for NLSST are1.01and0.60C;and for MCSST are0.48and1.16C.The matchups below4C come from colder regions of the Great Lakes,and are more susceptible to error caused by warm low clouds(fog)in the early morning during
AVHRR/SST accuracy for NOAA-12and-14satellites1299
Figure6.Scatter plots(satellite minus buoy SST measurements)vs air±sea temperature di V erence for the Gulf of Mexico,Northeast US and Southeast US coastal regions for1997.
We also plotted the bias of satellite minus buoy SST vs the wind velocity,air±sea temperature di V erence,satellite zenith angle and channel4and5brightness temperature di V erence for the Gulf of Mexico,Northeast US and Southeast US regions in gures5±8.We did not plot the similar gures for the Great Lakes region because the matchups for Great Lakes region are limited only to the summer.From gure5we can see that our SST matchup dataset covered the wind velocity range from0to17m s1.The SST bias was about the same in low wind situations as in
1300
X.L i et al.
Figure7.Scatter plots(satellite minus buoy SST measurements)vs satellite zenith angle for the Gulf of Mexico,Northeast US and Southeast US coastal regions for1997.
from16C to about+5C.Also,as the satellite zenith angle varied from0to about62,the SST bias did not change appreciably.These gures show that,for this set of parameters there was no single dominant factor that contributed to the SST bias.In the development of the NLSST algorithm(Walton et al.1998),it can be seen that the channel4and5temperature di V erence is correlated to the atmo-spheric absorption of infrared radiation.From gure8we can see that the NLSST gives good SST measurements over the entire range of channel4and5temperature di V erence.This demonstrates the universality of the NLSST algorithm over the range
AVHRR/SST accuracy for NOAA-12and-14satellites1301
Figure8.Scatter plots(satellite minus buoy SST measurements)vs AVHRR channel4and 5brightness temperature di V erence for the Gulf of Mexico,Northeast US and Southeast US coastal regions for1997.
6.Conclusion
The NOAA/NESDIS operational NLSST SST retrieval algorithm was validated using a matchup dataset of NOAA moored buoys and NOAA-12and NOAA-14 satellite measurements in three US coastal regions and the Great Lakes in1997.For the three US coastal regions,both NOAA-14daytime and night-time SST measure-ments had a bias less than0.2C with a standard deviation about1C.For NOAA-12,satellite measurements were also in good agreement with buoy measurements.
1302
X.L i et al.
For the Great Lakes region,we used the NLSST equation at night and MCSST equation during the day to calculate NOAA-14SST.The bias for both was about 0.4C,which was a little bit higher than that of the coastal areas.For NOAA-12, the linear MCSST algorithm gave good results for daytime SST with a bias of about 0.26C.Due to early morning fog in the summer in the Great Lakes region,the NOAA-12night-time algorithm yielded a fairly large SST bias.
The NLSST algorithm works well in all the study regions as well as under di V erent wind velocities,air±sea temperature di V erences,satellite zenith angles and AVHRR channel4and5temperature di V erences.Techniques for identifying early morning fog during the summer are required in order to improve night-time SST measurements for the Great Lakes.Simply substituting the MCSST equation for the NLSST equations improves the bias,but increases the standard deviation of satellite±buoy di V erence.
Acknowledgments
This research was funded by the NOAA/NESDIS Ocean Remote Sensing Program and the NOAA CoastWatch Program.
References
A nding,D.,and K auth,R.,1970,Estimation of sea-surface temperature from space.Remote
Sensing of Environment,1,217±220.
B arton,I.,1983,Dual channel satellite measurements of sea surface temperature.Quarterly
Journal of the Royal Meteorological Society,109,265±378.
B arton,I.,Z avody,A.,O’B rien,D.,
C utten,D.,S aunders,R.,and L lewell yn-J ones,D.,1989,
Theoretical algorithms for satellite-derived sea surface temperatures.Journal of Geophysical Research,94,3365±3375.
B arton,I.,1995,Satellite-derived sea surface temperature:current status.Journal of Geophysical
Research,100,8777±8790.
B o¨hm,E.,M arullo,S.,and S antoleri,R.,1991,AVHRR visible±IR detection of diurnal
warming events in the western Mediterranean Sea.International Journal of Remote Sensing,12,695±701.
B reaker,L.,K rasnopolsky,V.,R ao,D.B.,and Y an,X.H.,1994,The feasibility of estimating
ocean surface currents on an operational basis using satellite feature tracking methods.
Bulletin of the American Meteorological Society,75,2085±2095.
C ayula,J.F.,and C ornillon,P.,1992,Edge detection algorithm for SST images.Journal of
Atmospheric and Oceanic T echnology,9,67±80.
C ornillion,P.,and S tramma,L.,1985,The distribution of diurnal sea surface warming event
in the western Sargasso Sea.Journal of Geophysical Research,90,11811±11815.
C ornillion,P.,1989,Sea surface temperature products for the OceanographyScienti?c Research
Community.A report of the Sea Surface Temperature Archiving Science Working Group, sponsored by NASA and NOAA,Joint Oceanographic Institutions,Inc.,Washington, DC,36pp.
E mery,W.J.,and
F owler,C.W.,1991,Fram Strait satellite image-derived ice motion.Journal
of Geophysical Research,96,4751±4768.
E mery,W.J.,Y u,Y.,W ick,G.A.,S chluessel,P.,and R eynolds,R.W.,1994,Correcting
infrared satellite estimates of sea surface temperature for atmospheric water vapor attenu-ation.Journal of Geophysical Research,99,5219±5236.
H arries,J.E.,L lewell yn-J ones,D.T.,M innett,P.J.,S aunders,R.W.,and Z avody,A.M.,
1983,Observation of sea-surface temperature for climate research.Philosophical T ransactions of the Royal Society of L ondon Series A,309,381±395.
H awkins,J.D.,C lancy,R.M.,and P rice,J.F.,1993,Use of AVHRR data to verify a system
for forecasting diurnal sea surface temperature variability.International Journal of Remote Sensing,12,1347±1357.
K ahru,M.,H a·kansson B.,and R ud,O.,1995,Distributions of the sea-surface fronts in the
AVHRR/SST accuracy for NOAA-12and-14satellites1303 K rasnopolsky,V.M.,and B reaker,L.C.,1994,The problem of AVHRR image navigation revisited.International Journal of Remote Sensing,15,979±1008.
L lewell yn-J ones,D.T.,M innett,P.J.,S aunders,R.W.,and Z avody,A.M.,1984,Satellite multichannel infrared measurements of sea surface temperature of the N.E.Atlantic Ocean using AVHRR/2.Quarterly Journal of the Royal Meteorological Society,110, 613±631.
M ay,D.A.,and H ol yer,R.J.,1993,Sensitivity of satellite multichannel sea surface temperature retrievals to the air±sea temperature di V erence.Journal of Geophysical Research,98, 12567±12577.
M c C lain,E.P.,P ichel,W.G.,and W alton,C.C.,1985,Comparative performance of AVHRR-based multi-channel sea surface temperatures.Journal of Geophysical Research,90, 11587±11601.
M cmillin,L.M.,1975,Estimation of sea surface temperature from two infrared window measurements with di V erent absorption.Journal of Geophysical Research,80,5113±5117. M cmillin,L.M.,and C rosby,D.S.,1984,Theory and validation of the multi window sea surface temperature from space.Journal of Geophysical Research,89,3655±3661.
M innett,P.,1990,The region optimization of infrared measurements of sea surface temperature from space.Journal of Geophysical Research,95,13497±13510.
P earce,A.F.,P rata,A.J.,and M anning,C.R.,1989,Comparison of NOAA/AVHRR-2sea surface temperature with surface measurement in coastal waters.International Journal of Remote Sensing,10,37±52.
P ichel,W.G.,1991,Operational production of multichannel sea surface temperatures from NOAA polar satellite AVHRR data.Palaeogeography,Palaeoclimatology,Palaeoecology, 90,173±177.
P lanet,W.G.,1998,Data extraction and calibration of TIROS-N/NOAA radiometers.NOAA Technical Memorandum NESS107,Rev.1,US Department of Commerce,NOAA/ NESDIS,Washington,DC,105.
R obinson,I.S.,and W ard,N.,1989,Comparison between satellite and ship measurements of sea surface temperature in the north-east Atlantic ocean.International Journal of Remote Sensing,10,787±799.
S henoi,S.C.,1999,On the suitability of algorithms for the retrieval of SST from the north Indian Ocean using NOAA/AVHRR data.International Journal of Remote Sensing, 20,11±29.
S towe,L.L.,M c C lain,E.P.,C arey,R.,P ellegrino,P.,G utman,G.G.,D avis,P.,L ong,C., and H art,S.,1991,Global distribution of cloud cover derived from NOAA/AVHRR operational satellite data.Advanced Space Research,11,51±54.
S trong,A.E.,and M c C lain,E.P.,1984,Improved ocean surface temperatures from space-comparisons with drifting buoys.Bulletin of the American Meteorological Society,65, 138±142.
T opliss,B.J.,1995,A review of satellite sea surface temperature validations for NOAA’s7,9 and11using imagery o V eastern Canada.Canadian Journal of Remote Sensing,21, 492±510.
W alton,C.C.,1988,Nonlinear multichannel algorithms for estimating sea surface temperature with AVHRR satellite data.Journal of Applied Meteorology,27,115±124.
W alton,C.C.,P ichel,W.G.,S apper,J.F.,and M ay,D.A.,1998,The development and operational application of non-linear algorithms for the measurement of sea surface temperatures with the NOAA polar-orbiting environmental satellites.Journal of Geophysical Research,103,27999±28012.
Y ates,H.,C otter,D.,and O hring,G.,1985,ENVIROSAT-2000Report,Operational Satellite Support to Scienti?c Programs,NOAA,Washington DC,50.
Y okoy ama,R.,and T anba,S.,1991,Estimation of sea surface temperature via AVHRR of NOAA-9Dcomparison with xed buoy data.International Journal of Remote Sensing, 12,2513±2528.
Y okoy ama,R.,T anba,S.,and S ouma,S.,1993,Air±sea interacting e V ects to the sea surface temperature observation by NOAA/AVHRR.International Journal of Remote Sensing, 14,2631±2646.
海娜粉染发使用方法（自己总结，供参考） 1、染发前需先将头发洗净后并将头发晾干。（头发不脏就好）然后，选择一个没什么事、不出门的下午，在家请亲人帮忙向头发上涂抹，或者提前找好理发店请理发师傅帮忙。自己一个人没办法涂均匀的。 2、将海娜粉1袋100克放入一个大碗（饭店打包的塑料碗也可）中，加一些盐（就像腌咸菜那么多，帮助润渍、上色），将开的红茶水（最好泡20分钟之后用，更红，最好保持热度）冲调，用筷子搅拌，搅匀“很稠”的泥浆，先不要太稀，因为还要放鸡蛋。放稍凉一些，可以再打入一个鸡蛋（太热放鸡蛋，鸡蛋就熟了，呵呵），再用筷子调匀，比刚才就稀一些了。调匀后可放少许蜂蜜,一些维生素C（药店有售维生素C片，每次放10片左右）粉末，搅匀。要是觉得还是稠，再加些红茶水。最好是像稠的粥一样。不要太稠，那样不好涂抹；也不要太稀，那样会流的到处都是。 3、用细密的梳子，趁温热刷到头发上，太凉了涂在头上不舒服，怕感冒。先涂头顶，然后周围地方的头发都向头顶堆积。这个泥状的泥浆，比染发剂体积要多，涂时也有点不顺滑，只要保证白发的地方有泥，就能上色。没经验时，就尽量涂厚些。 4、在发际处缠一圈旧毛巾条或几圈卫生纸，防止留下来污染衣服。然后用保鲜膜将头发部分包紧，防止流下来污染。然后，就可以在家里做些事情。 5、4-6小时后，洗掉，保留时间越长，颜色越深。因为是花茎叶的泥，所以，洗时很费水。把泥冲少了，可以用一次洗发液，最好只用一次洗发液。等过几天再正常用多次洗发液。 海娜粉之外，其他作料的作用如下： 加点盐是为了海娜粉汁更浸透。 加鸡蛋一方面是为了营养，一方面是为了具有粘性。 加蜂蜜是为了头发更顺滑的作用。（可不加） 加橄榄油也是为了头发更顺滑一些。（可不加） 加红茶是为了颜色偏红。 加维C粉，目的是固色。（可不加） 海娜粉染发就是麻烦，但对身体还是好的。
中国信通院解读政务大数据标准化现状和趋势 日前，中国信息通信研究院云大所大数据技术主管姜春宇在“OSCAR云计算开源产业大会”上，围绕政务大数据标准化现状和趋势进行深入解读。中国信息通信研究院云大所大数据技术主管姜春宇大数据时代的到来给政府信息管理变革带来了新的契机 在大数据概念和技术出现之前，国家试图解决的是政务信息资源交换共享的问题，在2007年就推出了政务信息资源交换体系和政务信息目录的系列国家标准，从数据标准和交换体系方面试图解决数据交换共享的问题。随着大数据技术与应用的发展，政府面临新的任务： 一是利用大数据来提升政府决策和治理能力。除了实现政务信息的共享之外，还需要构建起大数据资源的汇集与整合，从而为政府各个部门提供完善的分析支撑的能力。 二是如何将政务的数据资源运营流转起来，对外辐射到整个社会各行各业，将价值释放给社会和民众，促进社会的发展进步，这就是数据分析应用和数据资产管理的需求。 国家大数据战略实施以来，我国政府出台了多项顶层设计，为大数据产业的快速成长提供良好的发展环境。特别是2017年起，'加快国务院部门和地方政府信息系统互联互通，形成全国统一政务服务平台'、'深入推进'互联网+'行动和国家大数
据战略'、等要求陆续提出，为政府信息化建设提供了新的商业机遇和建设方向。在多种因素的驱动下，国家和各地方政府围绕政务信息资源标准化发展，紧锣密鼓地发布了多个重要政策文件。其中，特别是《政务信息资源共享管理暂行办法》、《政务信息系统整合共享实施方案》、《政务信息资源目录编制指南（试行）》三个重要文件，不仅明确了政务信息资源共享的原则、分工，给出了信息系统整合的实施方案，也给出了国标《政务信息目录》标准体系正确打开方式，更具有实操性。这也说明国家认识到了标准的落地需要更多推广手段和指导手段。然而，我们也必须看到，当前在政务信息资源交换共享过程中，仍然在标准使用、业务系统建设、执行机构、数据共享全责等方面存在较多问题，需要进一步完善。 新挑战：政务数据治理和数据资产管理 政务数据资产管理是一个新的命题，在概念、目标与实施途径等方面，与传统的政务数据共享交换都存在差异。 大数据发展促进委员会发布的《数据资产管理实践白皮书》中，对数据资产管理的基本架构进行了描述，其中包含9个活动职能和2个保障措施。活动职能是指落实数据资产管理的一系列具体行为，保障措施是为了支持活动职能实现的一些辅助的组织架构和制度体系。 数据资产管理体系架构围绕这一体系，大数据发展促进委员
如何染发才不伤发？既不会伤害到头发，还能让白发瞬间变黑发！其实原料很简单，就是这几种：何首乌、姜、青黛、干松和白芨。”原来，青黛里面的靛青和乌发成分的何首乌这么一结合，黑色染发剂就初步成形了，再加上起到固色作用的白芨后，中药染发剂上色就更容易了。而干松和生姜，好处刚才人家杜医生也说了，涂抹在头发上既有芳香的味道，同时也能养发护发！配料有：姜，何首乌，青黛，干松、白芨，每样各20克。首先，把它们倒进搅拌机里，加入1升的水，打成汁！因为青黛本身就是粉状，所以不用加入豆浆机。您看，现在这几样东西已经被我们打磨成汁了！经过过滤后，上锅熬！您可别以为这个步骤很简单，专家告诉我们，选择什么样的锅也是有讲究的：中国中医科学院教授杜婕僡：“因为铁锅里面含有丰富的铁元素，它能让我们的头发吸收更多营养，还会让头发变得更加黑亮有光泽，所以应该选择用铁锅熬。”您记住了，得用铁锅。接下来就是火候的问题了，一般我们选择小火慢熬。因为只有熬成糊状，才能方便我们染发，所以得熬制一个小时左右才可以关火。对于用量方面，也是根据您的头发长短而定：一般像郭女士这样的中长发，每样东西大概用20克左右，如果是短发，差不多每样10克就足够了！接下来的步骤，就跟我们平时染发一样了！据郭女士说，用中药染过的头发，不容易掉色，也就是说，持续的时间比较长，坚持两三个月不成问题！首先说说“海娜粉”，是一剂内服药，它本身有活血化瘀，祛风止痛的作用，它不光能吃，染发效果也不错。造型师成远：“首先海娜粉本身是纯植物的，所以不言而喻，海娜粉在头发上涂抹的话很难一次性让它颜色到位，需要一次两
次多次地染，才能让头发有一个更好的变化。”[size=+0] 原来，海娜粉就是指甲花，它本身能染色，可因为它是纯植物配方，所以很难一次见效，一般要坚持四周，每天都要使用。虽然用海娜粉染色的时间比较长，不过专家告诉我们，它最大的优点就是在上色的同时，还能保护头发，而且能有效遮盖白色，并且，染出来的发色不是死板的黑色，而是微微有些发红，非常漂亮时尚，也显得人更加年轻。调制方法也很简单：海娜粉加上温水，调匀即可。海娜粉的用量是短发一次20——30克，中长发一次50——60克。另外，您还可以在其中加入少量蜂蜜和酸奶，能起到润发的作用。中国中医科学院教授杜婕僡：“杭白菊里面其实是不含有色素和其他褪色成分的，但是因为它里面含有氨基酸，它能够淡化我们头发的黑色素，也就是说它能褪去黑色，所以染发的效果也是很好的。”另外，杜医生还告诉我们，杭白菊里面含有丰富的维生素和微量元素，不仅染发效果明显，对头发还能起到非常好的保护作用。-看到这，估计好多人都有这么一想法：这费了半天劲染的头发，颜色要是说掉就掉了，多可惜呀！最后我再教您一招天然固色法：每天洗头的时候，您不妨放一些粗老的黑茶，因为黑茶里面除了含有丰富的营养物质外，还有固定色素的作用，无论您染了哪种颜色，黑茶都可以很好地固定颜色。这样一来，染发，护发的烦恼就全都解决了！ [size=+0] 下面是把白发染黑的做法： 材料：何首乌20g，白芨20g，青黛30g，干松20g，薄荷10g (记着请店员把药研碎)
《大数据技术原理与应用》课程标准 一、课程信息 课程名称：大数据技术原理与应用课程类型：考查课 课程代码：授课对象：物联网工程专业本科班，物联网创新班 学分：先修课：物联网导论、操作系统教程、编程 学时：后续课：智能家居、智能物流、云计算 制定人：理艳荣、张海兰制定时间： 二、课程性质 《大数据技术》是一门专业选修课，大数据技术入门课程，为学生搭建起通向“大数据知识空间”的桥梁和纽带，以“构建知识体系、阐明基本原理、引导初级实践、了解相关应用”为原则，为学生在大数据领域“深耕细作”奠定基础、指明方向。 课程将系统讲授大数据的基本概念、大数据处理架构、分布式文件系统、分布式数据库、数据库、云数据库、分布式并行编程模型、流计算、图计算、数据可视化以及大数据在互联网、生物医学和物流等各个领域的应用。在、、和等重要章节，安排了入门级的实践操作，让学生更好地学习和掌握大数据关键技术。 三、课程设计 1.课程目标设计 （1）能力目标 总体目标：通过学习大数据相关理论知识，掌握大数据的系统架构及关键技术以及具体应用场景，并结合具体设计实例，培养学生创新意识和实践能力。 件系统的重要概念、体系结构、存储原理和读写过程，并熟练掌握分布式文件系统的使用方法； （）能够了解分布式数据库的访问接口、数据模型、实现原理和运行机制，并熟练掌握的使用方法； （）能够了解数据库与传统的关系数据库的差异、数据库的四大类型以及数据库的三大基石；基本掌握、等数据库的使用方法； 具体目标：
序号单项能力目标 能够掌握大数据的基本概念 能够掌握相关的数据管理、存储、分析计算等的技术基础 能够掌握的相关知识 通过对数据库的学习和编程设计，掌握的使用方法 掌握大数据知识体系的系统架构 （2）知识目标 序号知识目标 了解分布式文件系统的基本概念、结构和设计需求，掌握的概念 了解布式数据库的访问接口、数据模型、实现原理和运行机制 了解数据库与传统的关系数据库的差异、数据库的四大类型以及数据库的三大基石 了解云数据库的概念、基本原理和代表性产品的使用方法 2.课程内容设计 （）设计的整体思路：面向实践，以理论知识与具体应用相结合的方式介绍大市聚。理 论结合实际，由浅入深，加强对大数据概念及技术的理解与巩固。此课程划分为下列模块。 （）模块设计表： 模块名称学时 介绍大数据的基本概念和应用领域，并阐述大数据、云计 算和物联网的相互关系 介绍大数据处理架构 分布式文件系统的基本原理和使用方法 分布式数据库的基本原理和使用方法 数据库的概念和基本原理 云数据库的概念和基本原理 分布式并行编程模型原理和使用方法 架构再探讨 总复习 合计 3.教学进度表设计
化妆品中7种染发剂的检测方法 （征求意见稿） 1 适用范围 本方法规定了化妆品中酸性紫43（CAS：224-618-7，CI 42510 ）；碱性紫14（CAS：632-99-5，CI42510）；酸性橙3（CAS：228-921-5）；碱性黄87（CAS：116844-55-4）；碱性蓝26（CAS：2580-56-5，CI44045）；碱性红51（CAS：12270-25-6）；碱性橙31（CAS：97404-02-9）7种染发剂含量的高效液相色谱测定方法。 本方法适用于染发类和烫发类化妆品中碱性红51、碱性紫14、酸性紫43、碱性蓝26、碱性橙31、酸性橙3、碱性黄87等7种染发剂含量的测定。 2方法提要 试样经甲醇超声提取后，过0.45μm滤膜，采用高效液相色谱系统分离，紫外检测器或二极管阵列检测器进行检测，外标法定量。若取5.0g样品，本方法对碱性红51、碱性紫14的检出限为0.06mg/kg；酸性紫43、碱性蓝26的检出限为0.6mg/kg；碱性橙31、酸性橙3的检出限为1.2mg/kg；碱性黄87的检出限为3.0mg/kg；最低定量浓度分别为：碱性红51、碱性紫14：0.2mg/kg、酸性紫43、碱性蓝26为2.0mg/kg；碱性橙31、酸性橙3为3.5mg/kg；碱性黄87为10mg/kg。 3试剂和材料 除另有规定外，所用试剂均为色谱纯，水为一级水。 3.1 甲醇。 3.2 四氢呋喃。 3.3 乙酸铵：分析纯。 3.4 乙酸铵溶液（0.02mol/L）：称取1.54 g乙酸铵（3.3），加水至1000 mL，溶解，经0.45μm滤膜过滤。 3.5 染发剂标准物质：酸性紫43（Acid Violet 43，CAS：224-618-7，CI 42510 ）；碱性紫14（Basic Violet 14，CAS：632-99-5，CI42510）；酸性橙3（Acid Orange 3，CAS：228-921-5）；碱性黄87（Basic yellow 87 ，CAS：116844-55-4）；碱性蓝26（Basic Blue 26，CAS：2580-56-5，CI44045）；碱性红51（Basic red 51，CAS：12270-25-6）；碱性橙31（Basic orange 31，CAS：97404-02-9）；纯度均≥90%。 3.6 染发剂混合标准储备液（1.0 mg/mL）：分别称取7种染发剂标准物质（3.5）0.1 g
植物染发剂使用方法 植物染发剂使用方法： 步骤一：用一玻璃制或塑料制品的碗,将等量的HerbatintHerbalHaircolorGel染发剂和GlycolDeveloper植物发 胶混合搅拌。 中短发：herbatint天然植物染发剂和乙二醇显色剂各30毫升; 长发：herbatint天然植物染发剂和乙二醇显色剂各60毫升。 视头发的长度,厚度,1盒可用1至2次,将混合液渗入头发,将瓶 盖盖回原瓶上。 步骤二：使用梳子,立刻将混合液涂抹于发上,将剩余未使用的混合液丢弃。 如之前染过发的:将混合液敷在新长出的头发上。等候30分钟, 将剩余的混合液涂抹于其它头发部位,包括发尾,再等10分钟(共计 40分钟)如之前未染过发(或染发的头发已经完全退色)：将混合液, 从发根到发尾,均匀地,局部地涂抹在头发上,不要梳头发。 等候40分钟。在染发过程中,你看到的颜色可能和你所选择的颜色有所不同,不用担心，这一般不会影响最终的染发效果. 步骤三：40分钟过后,用清水冲洗染发剂，直到水变清为止，然 后使用pH值为中性的洗发香波(建议使用herbatint修护洗发香波)，冲洗干净后擦干头发。 植物染发剂的分类： 1.植物染发剂在染发过程中不会产生刺鼻的异味。 2.植物染发剂不粘皮肤，不会造成染发后头发出现干枯、开叉。
海娜粉的种类 1.按颜色区分 原色，即红棕色，是海娜粉本来的颜色，也较基本色。目前国内的咖啡色、棕色、褐色、栗色等都是通过基本色这个颜色研制出来的。 2.特殊颜色 在使用的时候加入其它的植物，或者是加入少量的化学成分来进行调色，颜色丰富，上色效果明显增强，例如浅棕色、深棕色、酒红色、紫黑色等。 3.按产地划分 印度是海娜粉的主产国，目前国内大多数使用的大都产自巴基斯坦，但在国内分袋。我国新疆绝大多数的海娜粉就属于这一类，技术含量相对较低，加工方式简单，粉质较粗，所以价格较低，染发效果一般，颜色为单一的红棕色、黑色。 五倍子的分布地区 1.我国五倍子的主要产地集中分布于秦岭、大巴山、武当山、巫山、武陵山、峨眉山、大娄山、大凉山等山区和丘陵地带。
附件25： 高职电子信息大类大数据技术与应用赛项 技能竞赛规程、评分标准及选手须知 一、竞赛容 赛项名称：大数据技术与应用 赛项容：以大数据技术与应用为核心容，重点考察参赛选手在Hadoop平台环境下，对于大规模并行数据处理以及存计算技术的应用能力。具体包括： 1. 掌握Hadoop平台环境部署与基本配置，了解基于大数据计算平台的常见应用； 2．综合利用numpy、pandas、matplotlib、scikit 模块和MapReduce技术、分布式存储系统HDFS、分布式计算框架MapReduce/Yarn、数据仓库Hive、Python 等开发语言工具和技术，匹配和连接数据源，实现大数据的采集，提取、清洗、转换、分析、挖掘操作，产生分析结果，并且实现可视化呈现。 3．依据项目应用需求和分析结果，完成数据分析报告的编写。 二、竞赛方式 本赛项为团体赛，每支参赛队由3名参赛选手组成。 三、竞赛时量 竞赛时间4小时，竞赛连续进行。 四、名次确定办法 名次确定办法原则上按照竞赛总成绩从高分到低分排序确立选手名次。总成绩相同时，完成时间较短者名次列前；成绩和完成时间均相同时，操作过程较规者名次列前。 五、评分标准与评分细则 1．评分标准 本赛项总分为100分，采取分项计分制（表1）。
2．评分细则 竞赛成绩评定实行“裁判长合权负责制”，负责组织评分裁判进行成绩评定。评分裁判负责对参赛队伍（选手）的比赛作品、比赛表现按赛项评分标准进行评定。成绩评定根据竞赛考核目标、容对参赛队或选手在竞赛过程中的表现和最终成果做出评价。 本赛项的评分方法为现场评分和结果评分，现场评分为5分，由现场裁判根据参赛队的操作规以及综合表现情况进行评分；结果评分为95分，依据赛项评价标准，对参赛选手提交的竞赛成果进行评分。
（品名）L’OREAL Paris NEW excell10’ 保护头发抵御外来侵蚀 优化的PH值，对抗自由基。 镜光色泽，羊绒手感。 “除了色底10.0样式外，本品对其他色底均有效。如果你是轻微金发本色，本产品的色底难以完全覆盖您的白发。” 使用时按照本宣传页上所示各步程序和建议进行。为达理想结果，务必按照所指示的反应时间进行。
A. 一管戴帽的反应剂。 B. 一管色剂 C.一瓶带可可油的开司米香膏。 D.1付专业的高质量手套 E. 一个简易发梳涂抹器E和一个精准发际涂抹器F 快速准备 戴上手套D，把毛巾铺在肩膀周围以保护衣服。 只能使用反应剂A的帽来制备混合剂。 1．拧开反应剂A的瓶盖，打开色剂B（在前述皮肤过敏性测试中已经穿孔）。 把色剂B拧紧在反应剂A上。 2．把色剂B中所有的东西挤入反应剂A中。
---------------------------------------------------宣传页第二面------------------------------------------------- . 2．把色剂B 中所有的东西挤入反应剂A 中。 3．用反应剂A 原来的盖子将其小心盖好。 为取得最佳结果，详细阅读本插页并遵从相关反应时间，否则染发颜色可能暗淡。 选择适合你所要求的使用方法 快速简单使用 如果你还没选，先戴上手套D ，在肩膀周围围铺上毛巾以保护你的衣服。如果任何颜色掉到你的皮肤上，用湿布擦掉。 方法1 染过的头发 适于谁？此方法适用于：以前用同样色底染过发，而且发根正在长出。 如何用？先用于发根，等待反应，然后再用于长发和发梢，以避免已染的头发有过多染发剂。 用什么？针对于精确染发（染至发根），推荐使用精准发际涂抹器F 。 用之前：把头发弄湿，不用洗；然后用毛巾擦干，用梳子全面梳理。
第四章染发剂的选择及使用方法 第一节染发剂的分类 引言： 有什么样的染发剂可以提供给我们去选择呢？什么样的染发剂才适合顾客的实际情况呢？不同的发质怎样去选择适合发质本身的染发剂。. 一、非氧化染发剂 非氧化染发剂是不与显色剂混合使用的可直接涂到头发上的染发剂。这种染发剂只是在头发中沉淀颜色，产生物理变化，颜色会被一次次地洗掉。 1、临时性染发剂：顾名思义，临时性染发剂用于产生暂时性的颜色变化，并且会随着洗发次数的增加而逐渐褪色。它们是非反应性的直接使用的染发剂，也就是说，它们不需使用化学物来使之显色。它们含有较大的颜色分子，只能覆盖表皮的表面，因此只在头发中产生物理变化（而非化学变化）。产品中不含任何能漂浅头发的物质，配方或头发中也不发生任何化学变化。临时性染发剂不需要进行皮试或过敏测试，因为它们不含碱性衍生物，否则必须进行皮试。 临时性染发剂包括周染、彩色摩丝、彩色啫喱、彩色笔、彩色油、彩色喷发胶、彩色洗发水和冲洗剂等。与其他染发产品不同的是，暂时性染发剂涂到头发中之后，不须冲水。你只要将头发吹干并造型即可。 周染通常是在洗头盆中染发，用于为褪色的头发增加色调，中和掉不想要的色调或对未使用化学手段改变结构的头发暂时性增加色彩。 ⑴、彩色摩丝和彩色啫喱有多种颜色，用于使原有颜色更加鲜艳，为灰白发改善色调以及产生夸张效果。由于彩色摩丝也可增加发量，所以也在造型中使用。 ⑵、彩笔和彩油也有多种颜色，可用于产生各种效果，包括连接新旧发色，以及产生有趣的色彩缤纷的设计。 ⑶、彩色润发油的颜色范围也很广，除了包括为头发增加光泽，还能为头发增加色调或产生特殊的颜色效果。 ⑷、彩色喷发胶是气雾剂，密封罐装，属于易燃物质，所以不要在正在吸烟的人周围使用，也不要在明火周围使用。彩色喷发胶有许多种颜色，它能够快速而方便地在头发中建立颜色，产生特殊效果。 ⑸、染后洗发水和护发素用于染后头发的颜色维护，或增加色调，它们也可用来去除不想要的色调。 2、半永久性染发剂： 这种染发剂是碱性的，一般在洗数次头发后就褪色了，具体褪色情况还取决于头发的多孔性。它们含有大小两种颜 色分子，较小的颜色分子能够穿透头发的表皮层而进入皮质，而非象暂时性染发剂一样只能覆盖在头发表面。由于这些染发剂不使用化学方法来改变发色，所以它们只能沉淀颜色而不能漂浅头发，因此，根据所使用的颜色半永久性染发剂用来直接染发，不需混合，沉淀到头发中的颜色就是染发剂，这种染发剂在褪色的时候不会留下分界线，也不需要补染。 半永久性染发剂包括许多颜色，它们都被称为光泽剂或颜色加强剂，尽管这种染发剂不需与显色剂混合，但如果产 品中含有碱性衍生物成分，就必须先进行皮试。 根据产品中的成分以及头发的多孔性，重复使用半永久性染发剂会改变头发的结构，特别是
[参考论文]大数据存储技术标准化论文大数据存储技术标准化论文 摘要:大数据作为信息化时代的战略新兴产业，发展速度势不可挡，虽然目前国内还没有大数据存储的统一标准，但国内很多公司关注并投入到这一领域。制定符合中国国情的大数据存储接口标准，对促进整个产业的稳定发展具有重要的现实意义。 1 引言 随着互联网Web2.0的兴起和云计算的发展，大数据的价值越来越受到人们的重视，人们对数据的处理实时性和有效性要求也越来越高。大数据的应用已经进入了各行各业了，如商业智能、公共服务、科学研究等领域。目前大数据的分析技术发展十分迅速，尤其是大数据分析平台Hadoop得到了各大厂商的极大关注，基于Hadoop平台进行的大数据分析、数据存储研究正在进行[3]。目前国际、国内尚未出现大数据分析的全流程标准服务和接口定义，本文研究的重点是根据国内大数据的实际现状，采用hadoop平台进行大数据存储处理的全流程分析以及各个功能模块进行对比研究，提出建立大数据存储的标准化体系的建议，有利于促进形成大数据存储的基础性标准，从而为产业发展提供了有力的保障。 2 大数据存储技术的种类 大数据可能由TB级(或者甚至PB级)信息组成，既包括结构化数据(数据库、日志、SQL等)以及非结构化数据(社交媒体帖子、传感器、多媒体数据)[2]。大部分这些数据缺乏索引或者其他组织结构，可能由很多不同文件类型组成。针对不同类型的海量数据，业 界提出了不同的存储技术。 2.1 分布式文件系统
分布式文件系统主要代表有Google的GFS和Hadoop中的HDFS。GFS是一个可扩展的分布式文件系统，是针对与大规模数据处理和Google应用特性而设计的，他运行在廉价的普通硬件上，可以提供高容错、高性能的服务。 HDFS是开源的分布式文件系统(Hadoop Distributed File System)，运行在跨机架的集群机器之上，具有高吞吐量来访问大数据集应用程序。它采用了主/从结构，由一个NameNode节点和多个DataNode节点来组成，NameNode主节点是主服务器，管理文件系统的命名空间和客户端对文件的访问操作;DataNode是集群中一般节点，它负责节点的数据的存储。客户端通过NameNode 向DataNode节点交互访问文件系统，联系NameNode获得文件的元数，而文件I/O 操作则是直接和DataNode进行交互的。HDFS允许用户以文件的形式存储数据，HDFS将大规模数据分割成多个64M为单位的数据块，采用数据块序列的形式存储在多个数据节点组成的分布式集群中。它具有很强的可扩展性，通过在集群中增加数据节点来满足不断增长的数据规模，同时它也具有高可靠性和高容错性，每个数据块在不同的节点中有三个副本，在海量大数据处理方面有很强的性能优势。 2.2 半结构化数据NoSQL数据库 NoSQL是一种打破了关系型数据库长久以来占主导地位的快速成长起来的非关系松散数据存储类型，这种数据存储不需要事先设计好 的表结构，它也不会出现表之间的连接操作和水平分割。他可以弥补关系数据库在处理数据密集型应用方面表现出的性能差、扩展性差、灵活性差等问题，NoSQL数据库了是作为关系数据库的补充。目前主流的NoSQL数据库有文档型数据库、列存储数据库、键值对(Key-Value)存储数据库。 (1)列存储数据库:列式数据库是以列相关存储架构进行数据存储的数据库，主要适合与批量数据处理和即席查询[1]。列存储将所有记录中相同字段的数据聚合存储，它通常用于应付分布式存储文件系统。典型的列存储数据库有Cassandra、
教你如何自己在家染发10个步骤
教你如何自己在家染发10个步骤 想染发又嫌跑美发沙龙太麻烦？很多人都会想“自己动手丰衣足食”，但跟专业染发有专人侍候不一样，DIY染发是要抱不少风险的，挑错颜色、染错地方都可能让你的“杰作”上街时吓到别人。 1.忠于原色 无论你染前的发色，是天生自然发色还是已经人为改变过的发色，你挑选新发色的时候都必须忠于你的原色，根据你原来发色的色调来挑选。专业发型师建议无论是染深还是染浅，都不能和原来发色出入超过两个号的颜色，不然就会太夸张。 2.选用冷色调 如果你第一次染发，使用染浅发色的“暖色调”颜色，很容易就会把头发染得看起来缺乏光泽，像砂纸一样。所以要选择冷色调或中性色调的颜色，会安全很多。 3，染前做足储备 通常上一罐染发剂，足够给中长发人士使用一次（也就是头发到肩膀附近）。如果你的头发特别长，或者特别厚，就要在染发前做足储备，不然染到一半没有燃料就糗大了。 4.使用专业工具
通常上染发剂配备的染发用工具都比较简单，而且用途有限。所以要有更好的效果，就去美容工具店购买较为专业（起码较为结实）的工具。例如染发用手套就不能只是一层薄薄的透明塑啦。 5.先做颜色测试 与其染完之后不喜欢新发色而后悔莫及，就应该在染发之前，挑选一撮在表层以下的头发做颜色测试。 6.小心补染发根 DIY染发最容易出现的糗事就是补染发根颜色和原有发色不一致！就算使用同一个牌子同一个色号的染发剂，如果染剂在等待过程中从发根留向其他地方重复加色，就会导致补染和之前染得颜色不一样。要防止出现意外，你可以让朋友帮忙，或者找一块三面镜子来从各个角度看清楚下手途染剂的地方。如果你希望顺道也为已经染过的地方补一下颜色，就在快要清洗前的5分钟用梳子把染发剂顺梳到发梢，5分钟后全部头发一起清洗。 7.染后呵护发色常新 你越宝贝你的头发，它的颜色就会保持得越好。所以染发后挑选针对染色头发配制的洗护产品，至少每个星期做一次深层护理。 8.自然遮盖白发
大数据技术及应用-标准化文件发布号：（9456-EUATWK-MWUB-WUNN-INNUL-DDQTY-KII
大数据技术及应用 随着互联网技术的飞速发展，特别是近年来云计算、物联网、社交网络等新兴服务促使人类社会的数据种类和规模正以前所未有的速度增长,大数据时代正式到来.数据从简单的处理对象开始转变为一种基础性资源,如何更好地管理和利用大数据已经成为普遍关注的话题. 下面对大数据技术的应用做一些简单说明 （一）行业拓展者，打造大数据行业基石 IBM大数据提供的服务包括数据分析，文本分析，蓝色云杉（混搭供电合作的网络平台）;业务事件处理；IBM Mashup Center 的计量、监测和商业化服务（MMMS）。 IBM的大数据产品组合中的最新系列产品的InfoSphere bigInsights，基于Apache Hadoop。该产品组合包括：打包的Apache Hadoop的软件和服务，代号是bi gInsights核心，用于开始大数据分析。软件被称为bigsheet，软件目的是帮助从大量数据中轻松、简单、直观的提取、批注相关信息为金融，风险管理，媒体和娱乐等行业量身定做的行业解决方案微软：2011年1月与惠普（具体而言是HP数据库综合应用部门）合作目标是开发了一系列能够提升生产力和提高决策速度的设备。 EMC：EMC 斩获了纽交所和Nasdaq；大数据解决方案已包括40多个产品。 Oracle：Oracle大数据机与Oracle Exalogic中间件云服务器、Oracle Exadata数据库云服务器以及Oracle Exalytics商务
智能云服务器一起组成了甲骨文最广泛、高度集成化系统产品组合。 （二）大数据促进了政府职能变革 重视应用大数据技术，盘活各地云计算中心资产：把原来大规模投资产业园、物联网产业园从政绩工程，改造成智慧工程；在安防领域，应用大数据技术，提高应急处置能力和安全防范能力；在民生领域，应用大数据技术，提升服务能力和运作效率，以及个性化的服务，比如医疗、卫生、教育等部门；解决在金融，电信领域等中数据分析的问题：一直得到得极大的重视，但受困于存储能力和计算能力的限制，只局限在交易数型数据的统计分析。一方面大数据的应用促进了政府职能变革，另一方面政府投入将形成示范效应，大大推动大数据的发展。 （三）打造“智慧城市” 美国奥巴马政府在白宫网站发布《大数据研究和发展倡议》，提出“通过收集、处理庞大而复杂的数据信息，从中获得知识和洞见，提升能力，加快科学、工程领域的创新步伐，强化美国国土安全，转变教育和学习模式”；中国工程院院士邬贺铨说道，“智慧城市是使用智能计算技术使得城市的关键基础设施的组成和服务更智能、互联和有效，随着智慧城市的建设，社会将步入“大数据”时代。” （四）舆情分析
如何染发才不伤发？ 既不会伤害到头发，还能让白发瞬间变黑发！ 其实原料很简单，就是这几种：何首乌、姜、青黛、干松和白芨。” 原来，青黛里面的靛青和乌发成分的何首乌这么一结合，黑色染发剂就初步成形了，再加上起到固色作用的白芨后，中药染发剂上色就更容易了。而干松和生姜，好处刚才人家杜医生也说了，涂抹在头发上既有芳香的味道，同时也能养发护发！ 配料有：姜，何首乌，青黛，干松、白芨，每样各20克。- 首先，把它们倒进搅拌机里，加入1升的水，打成汁！因为青黛本身就是粉状，所以不用加入豆浆机。您看，现在这几样东西已经被我们打磨成汁了！ 经过过滤后，上锅熬！您可别以为这个步骤很简单，专家告诉我们，选择什么样的锅也是有讲究的： 中国中医科学院教授杜婕僡：“因为铁锅里面含有丰富的铁元素，它能让我们的头发吸收更多营养，还会让头发变得更加黑亮有光泽，所以应该选择用铁锅熬。” 您记住了，得用铁锅。接下来就是火候的问题了，一般我们选择小火慢熬。因为只有熬成糊状，才能方便我们染发，所以得熬制一个小时左右才可以关火。
对于用量方面，也是根据您的头发长短而定：一般像郭女士这样的中长发，每样东西大概用20克左右，如果是短发，差不多每样10克就足够了！ 接下来的步骤，就跟我们平时染发一样了！据郭女士说，用中药染过的头发，不容易掉色，也就是说，持续的时间比较长，坚持两三个月不成问题！ 首先说说“海娜粉”，是一剂内服药，它本身有活血化瘀，祛风止痛的作用，它不光能吃，染发效果也不错。 造型师成远：“首先海娜粉本身是纯植物的，所以不言而喻，海娜粉在头发上涂抹的话很难一次性让它颜色到位，需要一次两次多次地染，才能让头发有一个更好的变化。”- [size=+0]原来，海娜粉就是指甲花，它本身能染色，可因为它是纯植物配方，所以很难一次见效，一般要坚持四周，每天都要使用。虽然用海娜粉染色的时间比较长，不过专家告诉我们，它最大的优点就是在上色的同时，还能保护头发，而且能有效遮盖白色，并且，染出来的发色不是死板的黑色，而是微微有些发红，非常漂亮时尚，也显得人更加年轻。 调制方法也很简单：海娜粉加上温水，调匀即可。海娜粉的用量是短发一次20——30克，中长发一次50——60克。另外，您还可以在其中加入少量蜂蜜和酸奶，能起到润发的作用。
大数据条件下档案规范化和标准化建设【摘要】随着大数据时代的到来，对传统的档案管理工作提出了更高要求，当前档案部门在档案管理工作推进过程中，一定要注重档案规范化和标准化建设，主动适应大数据时代的发展趋势。目前，在档案规范化和标准化建设过程中还存在诸多问题，因此我们有必要以问题为导向，进一步提升档案规范化和标准化意识，打造标准化档案网站，更好地服务经济社会发展。 【关键词】大数据;档案规范化和标准化;存在的问题;优化路径 大数据时代的到来对人们的生产生活都产生了深刻影响，同时也为档案管理工作带来了新的发展契机，大数据的运用有利于提升档案规范化和标准化建设水平，但同时我们也看到当前部分档案管理部门对于大数据重视程度不足，在具体应用过程中相关制度体系建设不够完备，缺乏专业的大数据档案管理人才，这些问题都在一定程度上制约了档案规范化和标准化建设。 一、大数据时代对档案管理提出的新要求 大数据时代，数字档案数量不断增加，经济社会发展对档案服务水平提出了更高要求，同时档案部门也加强了档案数据化建设，这就对当前的档案管理工作提出了新要求。对于档案管理工作人员而言，不能停留在传统的管理理念之上，应该主动适应大数据时代的发展趋势，更新大数据管理理念，树立档案“大数据”观，提升“大服务”能力，构建和完善档案大数据平台。 （一）要树立档案“大数据”观。当前在档案管理过程中迫于人力、物力和财力的限制，很多档案部门仍然沿用传统的档案管理方法，对于档案馆的大量馆藏资源没有进行数字化转换，致使很多具有一定信息价值的档案资源仍然存放在库房，无法发挥其应有作用，相当一部分档案数据在历史的长河中不断被遗忘。在这一背景之下，我们必须树立档案“大数据”观，应该将档案馆的所有馆藏资源视为档案部门的大数据资源，并按照这些思路积极向上级部门和地方政府争取资金和人才支持，同时要向主管领导宣传和讲解档案大数据观的意义和价值，争取主管领导的支持。 （二）要提升“大服务”能力。在大数据观背景下，需要对所有的档案资源进行数字化转换，并运用大数据信息技术提升档案部门的服务能力。档案管理
啊酷奴染发膏使用方法： 1.将一袋染发膏挤到容器里，充分搅拌成水状（短发1袋，长发2-3袋，咖啡膏用量多一些） 2.将染发膏原液均匀涂抹在头发上揉5分钟，静置10分钟（咖啡色静置时加热帽加热） 3.第二次揉搓出充分的泡沫5分钟（确保上色均匀），如果泡沫少了可喷少许热水揉搓，再静置20分钟（咖啡色静置时加热帽加热） 4.待头发上色后，冲洗干净 5.再用护发素或固色膏揉洗3分钟，使毛鳞片闭合 6.用吹风机充分将头发吹干！ 注意事项： 1.带上手套（因指甲和头发的材质相同，染发膏虽不染皮肤却会染到指甲） 2. 不要兑水，也不要弄湿头发！全程不要使用其他洗发水（啊酷奴染发膏洗发护发染发三效合一） 3.冬天天气冷，染发膏在打开包装前用温热水温热一下 4.摩丝，发油，发蜡，弹力素免洗精华等产品需要事先洗掉吹干后再用啊酷奴染发膏！ 5.如果发质很硬并且有很多白发想染咖啡色，请在咖啡膏里适当加黑发膏（1/2袋咖啡膏加10滴黑发膏，剩余的黑发膏立刻用火封上待下次再用）搅拌均匀抹在白发处，停留10分钟再用咖啡膏均匀涂在头发上 6.过敏者请事先在耳后做测敏24小时 注：一盒6袋，全国统一零售价128元。 黑发膏竹炭纤维黑素提取，咖啡膏是咖啡豆萃取的咖啡因子配方，物理染发PK传统化学染发，健康！！啊酷奴染发膏使用方法： 1.将一袋染发膏挤到容器里，充分搅拌成水状 （短发1袋，长发2-3袋，咖啡膏用量多一些） 2.将染发膏原液均匀涂抹在头发上揉5分钟，静 置10分钟（咖啡色静置时加热帽加热） 3.第二次揉搓出充分的泡沫5分钟（确保上色均 匀），如果泡沫少了可喷少许热水揉搓，再静置 20分钟（咖啡色静置时加热帽加热） 4.待头发上色后，冲洗干净 5.再用护发素或固色膏揉洗3分钟，使毛鳞片闭 合 6.用吹风机充分将头发吹干！ 注意事项： 1.带上手套（因指甲和头发的材质相同，染发膏 虽不染皮肤却会染到指甲） 2. 不要兑水，也不要弄湿头发！全程不要使用 其他洗发水（啊酷奴染发膏洗发护发染发三效合 一） 3.冬天天气冷，染发膏在打开包装前用温热水温 热一下 4.摩丝，发油，发蜡，弹力素免洗精华等产品需 要事先洗掉吹干后再用啊酷奴染发膏！ 5.如果发质很硬并且有很多白发想染咖啡色，请 在咖啡膏里适当加黑发膏（1/2袋咖啡膏加10 滴黑发膏，剩余的黑发膏立刻用火封上待下次再 用）搅拌均匀抹在白发处，停留10分钟再用咖 啡膏均匀涂在头发上 6.过敏者请事先在耳后做测敏24小时 注：一盒6袋，全国统一零售价128元。 黑发膏竹炭纤维黑素提取，咖啡膏是咖啡豆萃取 的咖啡因子配方，物理染发PK传统化学染发， 健康！！ 啊酷奴染发膏使用方法： 1.将一袋染发膏挤到容器里，充分搅拌成水状 （短发1袋，长发2-3袋，咖啡膏用量多一些） 2.将染发膏原液均匀涂抹在头发上揉5分钟，静 置10分钟（咖啡色静置时加热帽加热） 3.第二次揉搓出充分的泡沫5分钟（确保上色均 匀），如果泡沫少了可喷少许热水揉搓，再静置 20分钟（咖啡色静置时加热帽加热） 4.待头发上色后，冲洗干净 5.再用护发素或固色膏揉洗3分钟，使毛鳞片闭 合 6.用吹风机充分将头发吹干！ 注意事项： 1.带上手套（因指甲和头发的材质相同，染发膏 虽不染皮肤却会染到指甲） 2. 不要兑水，也不要弄湿头发！全程不要使用 其他洗发水（啊酷奴染发膏洗发护发染发三效合 一） 3.冬天天气冷，染发膏在打开包装前用温热水温 热一下 4.摩丝，发油，发蜡，弹力素免洗精华等产品需 要事先洗掉吹干后再用啊酷奴染发膏！ 5.如果发质很硬并且有很多白发想染咖啡色，请 在咖啡膏里适当加黑发膏（1/2袋咖啡膏加10 滴黑发膏，剩余的黑发膏立刻用火封上待下次再 用）搅拌均匀抹在白发处，停留10分钟再用咖 啡膏均匀涂在头发上 6.过敏者请事先在耳后做测敏24小时 注：一盒6袋，全国统一零售价128元。 黑发膏竹炭纤维黑素提取，咖啡膏是咖啡豆萃取 的咖啡因子配方，物理染发PK传统化学染发， 健康！！
陶博士详细说明书 尊敬的顾客：一生好！良心品质，一切为您！ 本品为天然植物配方，不含化学成分，故不需要做过敏测试。植物配方的染发剂相对化学染发剂来说，操作比较麻烦，上色较慢，且不易渗透，故黑发效果比传统化学染发剂稍差，但是非常的安全健康，即使经常使用，也没有任何毒副作用，以及致癌，至畸的任何潜在危险。为保证染发效果，使用前请认真阅读此说明书和注意事项。 操作前准备： 使用非焗油型洗发水清洁头发表层的油污（记住，一定要是不含焗油成分的洗发水哦。因为焗油成分会附在发丝表面形成保护膜，导致黑色植物精华比较难渗透），再用毛巾擦干至6-7成干，这样有利于植物黑发素的渗透，提高黑发效果。 两步操作： 第一步：用梳将第一剂（1剂植物黑发精华）膏体
涂至头发，根据头发长短决定用量，用梳子正反方向反复梳理均匀，戴上浴帽用加热帽加热10分钟后关掉开关保温状态加热20-30分钟。对多年没染发的纯白干枯发质，可适当延长加热时间。然后用清水简单清洗，用毛巾擦至不滴水（注意：在清水的时候不要清洗过度，稍微用水简单清洗一下头发的表皮层即可，不要用洗发水洗，以免把刚刚渗透到头发里的第一剂都洗掉）。 第二步：操作方法同第一步一样，将第二剂（2剂植物黑发伴侣）用梳子充分梳理均匀，戴上浴帽用加热帽加热10分钟后关掉开关保温状态加热20-30分钟。对多年没染发的纯白干枯发质，可适当延长加热保温时间。 加热完毕后也可以不用洗发水冲洗，因为天然的植物精华对头发有焗油和护理效果。 注意事项： 1.膏体上到头发上后，一定要反复的梳理，使得毛鳞片打开，植物精华更易渗透进去。 2.加热保温温度30-40℃为最佳，保温时间越长黑发效果越好。（视发质而定，粗硬发质比较难渗
附件25 ： 高职电子信息大类大数据技术与应用赛项 技能竞赛规程、评分标准及选手须知 一、竞赛内容 赛项名称：大数据技术与应用 赛项内容：以大数据技术与应用为核心内容，重点考察参赛选手在Hadoop 平台环境下，对于大规模并行数据处理以及内存计算技术的应用能力。具体包括： 1. 掌握Hadoop 平台环境部署与基本配置，了解基于大数据计算平台的常见应用； 2 ．综合利用numpy 、pandas 、matplotlib 、scikit 模块和MapReduce 技术、分布式存储系统HDFS 、分布式计算框架MapReduce/Yarn 、数据仓库Hive 、Python 等开发语言工具和技术，匹配和连接数据源，实现大数据的采集，提取、清洗、转换、分析、挖掘操作，产生分析结果，并且实现可视化呈现。 3．依据项目应用需求和分析结果，完成数据分析报告的编写。 二、竞赛方式 本赛项为团体赛，每支参赛队由 3 名参赛选手组成。 三、竞赛时量 竞赛时间 4 小时，竞赛连续进行。 四、名次确定办法名次确定办法原则上按照竞赛总成绩从高分到低分排序确立选手名次。总成绩相同时，完成时间较短者名次列前；成绩和完成时间均相同时，操作过程较规范者名次列前。 五、评分标准与评分细则 1．评分标准 本赛项总分为100 分，采取分项计分制（表1）。
表1考核环节及评分标准 2 ?评分细则 竞赛成绩评定实行“裁判长合权负责制”，负责组织评分裁判进行成绩评定。评分裁判负责对参赛队伍（选手）的比赛作品、比赛表现按赛项评分标准进行评定。成绩评定根据竞赛考核目标、内容对参赛队或选手在竞赛过程中的表现和最终成果做出评价。 本赛项的评分方法为现场评分和结果评分，现场评分为5分，由现场裁判根据参赛队的操作规范以及综合表现情况进行评分；结果评分为95分，依据赛项评价标准，对参赛选手提交的竞赛成果进行评分。
植物染发剂的使用方法 很多中老年人在染头发的时候都会使用到植物染发剂。植物染发剂对皮肤的伤害是特别小的，一般情况下是需要把头发均和染发，还有就是大家也要注意做好头发的保湿工作，而很多中老年人是不知道植物染发剂的使用方式的，今天我们就向大家做详细的介绍。 1)染发前需先将头发洗净后并将头发的水迹擦干。 2)用开水、热红茶、热咖啡更佳搅匀调泥浆。或者直接用70度左右的水冲海娜粉。 3)或者可以再打入一个鸡蛋调匀后放少许蜂蜜,少许食盐,搅匀。然后放置30分钟待用，最好放置1小时以上(如果想使头发色泽加深，将用海娜粉的泥浆糊糊放置1小时后再敷用)，或者放置一个晚上，隔夜用效果更佳。(鸡蛋蜂蜜这些东西也可以不要加，避免自己操作到处流，弄脏衣服。) 4)如果感觉稠,可以再放点茶水,不能太稀哦,宁可稠一点也不能太稀,否则刷到头发上要流下来的。 5)刷时可用牙刷,用小排刷最好,把头低下来头发垂下去,在肩上铺上一张塑料,注意要一层一层刷, 从发根处开始刷,挑起一层头发刷好后放在肩上,再挑起一层继续刷,这样把整个头发刷完,然后用浴帽或塑料袋包起来。 6)海娜粉(或海娜膏)停留置在头发上的时间，视每人发质而定。自然温度下，可停留4小时，停留时间越长，颜色越深，如使用加热帽，加热20分钟，再停留约2-6小时，待头发冷却为止。 7)4-6小时后，将头发用温水将染发泥冲洗干净，头发呈棕色，6-12小时内呈深棕色或深褐色、酒红色，注意染发后不要立即使用洗发水洗头，以免影响染发效果，使用护发素即可。以上就是我们向大家介绍的植物染发剂的使用方法了。在平常生活当中，如果使用植物染发剂来染发，我们建议大家可以根据上面介绍的步骤去染发，但是大家也不要经常染发，因为这毕竟会对头发以及头皮健康有一定的伤害。