Documentation of *.swf files: The .swf files contain results of Chuck Long's clear-sky detection, fitting, and cloud effect analyses. Before using these data, the user should refer to the following publication for descriptions of the products and methods used. Long, C. N. and T. P. Ackerman, (2000): Identification of Clear Skies from Broadband Pyranometer Measurements and Calculation of Downwelling Shortwave Cloud Effects, JGR, (July or August, 2000). The .swf files are written in ASCII (text) and may be read using a free format. Below are descriptions of terms in the header records of the .swf files. The variable names in the left column (Date, DFRa, etc.) appear as they do in the header records. To the right of the variable name is an explanation of the variable and its units. The calculations are presented for each three minute averaging period that SURFRAD data are reported. Each .swf file has two sections. The first section has a header record followed by one line of data that lists the coefficients of the best-fit equations appropriate for the day the file represents. The header record contains the coefficient variable names and beneath these, in the second record, are the coefficients. The variable names for the coefficients are listed below in the left column under the heading 'Coefficient Data:' The best fit equations described by the coefficients are of the form: c + a(cosZ)^b, where cosZ is the cosine of the solar zenith angle (the independent variable for all best fit equations), c is the offset, a is the multiplier, and b is the exponent to which cosZ is raised. If the c coefficient is not given, its value is zero (i.e. no offset). Coefficient Data: Date - date in YYYYMMDD format, based on LST DFRa - "a" coefficient calculated for diffuse ratio fit (diffuse/global SW) DFRb - "b" coefficient calculated for diffuse ratio fit CSWa - "a" coefficient calculated for clear-sky global SW fit CSWb - "b" coefficient calculated for clear-sky global SW fit CSSWa - "a" coefficient calculated for clear-sky component sum SW fit (dir + dif) CSSWb - "b" coefficient calculated for clear-sky component sum SW fit (dir + dif) SCORc - offset coefficient calculated for clear-sky global-sum difference fit SCORa - "a" coefficient calculated for clear-sky global-sum difference fit SCORb - "b" coefficient calculated for clear-sky global-sum difference fit (These best-fit calculations are added to TSW to get estimate SSW, when SSW is not possible owing to insufficient information, e. g., no direct or diffuse measurements) ALBa/UVBa/PARa - "a" coefficient calculated for clear-sky albedo, UVB, and PAR fits ALBb/UVBb/PARb - "b" coefficient calculated for clear-sky albedo, UVB, and PAR fits SWratC - offset coefficient for the best fit correction equation to estimate replacements for missing SSW data SWRatA - "a" coefficient calculated for the best fit correction equation to estimate replacements for missing SSW data SWRatB - "b" coefficient calculated for the best fit correction equation to estimate replacements for missing SSW data The result of the SWrat equation is multiplied times CSW to give an estimate for SSW when there is not enough information to get a component sum. The second section contains the three-minute data and calculated variables. The third line of the file is the header record for the second section and contains the variable names of the data that follow. These variable names are listed below in the left column under the heading 'Irradiance data:' To the right of the variable name is an explanation of the variable and its units. In these descriptions, ÒmeasuredÓ refers to measurements and clear-sky refers to estimated clear-sky dataÑor what would have been measured had no clouds been present in the hemispheric sky. The time on each line represents the ending time of the averaging period. The three flags (Tflg, dflg, and rflg) are quality control flags that were derived from combinations of the clear-sky fit results. Clrf is a flag that indicates whether the three minute period was determined to be totally cloud free, based on the irradiance data. Irradiance data: Zdate - date in YYYYMMDD format, based on GMT Ztim - time in hhmm format, based on GMT Ldate - date in YYYYMMDD format, based on LST Ltim - time in hhmm format, based on LST CosZ - Cosine of the solar zenith angle (used for fitting) AU - earth-sun distance in AUs tsw - measured downwelling SW from global pyranometer (W/m^2) csw - clear-sky downwelling SW from global pyranometer (W/m^2) tswfcg - csw-tsw, or total short wave cloud effect as measured from the global pyranometer (W/m^2) dif - measured downwelling diffuse SW (W/m^2) cdif - clear-sky downwelling diffuse SW (W/m^2) diffcg - cdif-dif of diffuse cloud effect (W/m^2) dir - measured downwelling direct SW (W/m^2) cdir - clear-sky downwelling direct SW (W/m^2) difr - measured diffuse ratio (dif/tsw) cdifr - clear-sky diffuse ratio (cdif/csw) ssw - measured downwelling component sum (dif+dir) SW (W/m^2) cssw - clear-sky downwelling component sum (cdif+cdir) SW (W/m^2) sswfcg - cssw-ssw, or short wave cloud effect computed from the component sum (W/m^2) sflg - "sum flag" - for Sum data and other QC 0 = sum OK 1 = estimated from TSW plus correction function (orig. bad) -1 = bad diffuse or direct, and couldn't estimate -2 = CosZ > 0.25, |TSW/CSW - SSW/CSSW| > swrlim -3 = CosZ > 0.09, |TSW/CSW - SSW/CSSW| > 2*swrlim -4 = CosZ > 0.12, TSW/CSW > 0.9 AND Dif/TSW > 0.9 The "-2" and "-3" tests detect when one of either the direct or diffuse is "bad data". The "-4" tests detect when the whole tracker was off, i.e. the Sum is about equal to TSW because the Diffuse is unshaded. Tflg - "total SW flag" - for TSW 0 = OK 1 = TSW < -20.0 W/m^2 (bad data) 2 = TSW > 1500 W/m^2 (bad data) dflg - "diffuse flag" - for diffuse 0 = OK 1 = Dif < -20.0 W/m^2 (bad data) 2 = absolute difference between TSW and Sum > 100.0 W/m^2 (bad) 9 = diffuse estimated from TSW-Dir rflg - "direct flag" - for direct irradiance 0 = OK (measured) 1 = original data bad, estimated by TSW-Dif -9 = bad or missing, can't be estimated clrf - clear detection flag 0 = not detected as clear-sky 1 = this measurement detected as clear-sky and used for fitting alb - measured albedo (no units) calb - clear-sky albedo (no units) PAR - measured PAR (W/m^2) CPAR - clear-sky PAR (W/m^2) PARFcg - PAR cloud effect (W/m^2 UVB - measured UVB (milliW/m^2) CUVB - clear-sky UVB (milliW/m^2) UVBFcg - UVB cloud effect (milliW/m^2) Sswrat - ratio of the std. dev. of global pyranometer measurement divided by CSW (used to distinguish overcast from partly cloudy skies)