GEOPHYSICAL RESEARCH LETTERS, VOL. 23, NO. 16, PAGES 2121-2124, AUGUST 1, 1996 New ultraviolet spectroradiometer measurements at Mauna Loa Observatory B. A. Bodhaine, R. L. McKenzie, P. V. Johnston, D. J.Hofmann, E.G. Dutton, R. C. Schnell, 3J.E. Barnes, 3S.C. Ry .an, 3and M. Kotkamp Abstract, A research-grade scanning UV spectroradiometer wasinstalled at MaunaLoa Observatory (MLO), Hawaii, in July 1995. This instrument, built around a commercially available double monochromator, is interfaced with a PC to provide auto- maticcontrol anddataacquisition. The spectral range sampled by the instrument is 290-450 nm, and ,the bandpass is about1 nm. A complete scan requires about 200 seconds and is per- formed every 5 degrees of solar zenith angle (SZA) during day- light hours. Calibration is performed On site at 6-month intervals using a 1000-W standard quartz-halogen FEL lamp with calibration traceable to NIST. The •UV irradiances meas- ured at MLO are much more intense than at low altitude mid- latitude •locations. For observations at a SZA of 45 ø, the erythemally weighted UVcan ,excee d t8 •tW cm -2, which isap- proximately 15-20% greater than the maxima seen at Laude,r, NeW Zealand, forsimilar ozone amounts. The difference is pri- marily due to the higheraltitudeat MLO. For overhead sun conditions at MLO, erythemal UV can exceed 45 gWcm -2, which to our knowledge is the highest recorded •anywhere at the Earth's surface. UV irradiance is strongly correlated (inversely) with Dobsonspectrophotometer total ozonemeasurements at MLO, with highercorrelations at shorter wavelengths. The ra- diative amplification factor (RAP) for erythema at MLO is about 1.44 ñ 0.46 at SZA 45 ø. Using ozone retrievals from the UV spectra themselves, the deduced RAF for erythema is 1.26 ß 0.38. The RAPs for e,rythema at SZA 60 ø are similar, and in agreement with other determinations withinthe limitsof experi- mental uncertainty. Introduction A research-grade ultraviolet (UV) spectroradiometer was in- stalled at MaunaLoa Observatory (lVlLO), Hawaii, in July1995. MLO, one of the premier atmospheric monitoring sites in the world, is located on the island of Hawaii at an altitude of 3.4 km. Because Mauna Loa (mountain) extends above the marine boundary layer, and because of a diurnal upslope-downslope wind circulation, morning skiesat MLO are often unusually clear, providing excellent conditions for solar radiation measure- ments. MLO is operated by the Climatic Monitoring and Diag- nosticsLaboratory (CMDL) of the National Oceanic and Atmospheric Administration (NO•). The coordinates of the site are19.53 N, 155.58 W. tClimate Monitoring and Diagnostics Laboratory, National Oceanic and Atmospheric Administration, Bould•, Colorado •National Imtimte ofWater and Atmospheric Research, Lauder, Central Ota•, New Zealand aunaLoa Observatory, NationalOceanic and Atmospheric Admini- stration, Hilo, Hawaii Copyright 1996 by theAmerican Geophysical Union. Paper number 96GL01954 0094-8534/96/96GL-01954505.00 The instrument described here was developed and operated bythe National Institute forWater and Atmospheric Research (NIWA)at Lauder, NewZealand, and has been included in a number of spec,troradiometer intercomparisons [McKenzie et al., 1991; McKenzie etal., 1993; Seckmeyer etal., 1995]. The solar radiation measured at the earth's surface depends on the absorption and scattering of the atmosphere, the earth-sun distance, and the irradiance ofthe sun.Thespectral solar irradi- ance above the earth's atmosphere was presented by Nicolet [1989]. The atmospheric transmission in the UV •portion of the spectnm!. is controlled primarily by Rayleigh scattering by air molecules, scattering by clouds, and absorption byozone. Since ozone is affected by anthropogenic influences, solar UV irradi- ance arriving at theearth's surface is controlled by both natural and anthroP0.genic effects. Ozone concentration, in turn, is also affected bychanges in solar UV. The UV-Aregion ofthe •spec- tram (320-400 nm) is virtually unaffected by ozone absorption; the UV2B (280-320 nm) is Strongly affected byvariations in ozone; and the UV-C (< 280 nm) isalmost entirely ab•ørbed be- fore it reaches thesurface. An excellent review of this subject was provided byStamnes [1993]. Instrumentation The UV spectroradiometer is built around a commercially available Jobin-Yvon DH10 double monochromator and is inter- faced with a computer to provide automatic control anddata ac- quisition [McKenzie etal., 1992].The spectral range sampled by the instrument is 290-450 nm, andthe bandpass is abou•t 1 nm. Thegratings aredrip'yen •y a stepper motor under computer control, the irradiance signal is sampled every0.2 •nm, and a complete scan requires about200 seconds. The instnm•.ent is mounted in an insulated weatherproof enclosure (painted white) located ona concrete pad at the MLO site. The interior of the enclosure is temperature controlled to 20 ñ 0.5 C using a Peltier heater/cooler unit.A machined quartz dome with active aircir- culation to inhibitcondensation protects the diffuser fromthe weather. Calibration of the spectroradiometer is performed on site at 6-month intervals using a standard 1000-W FEL quartz-halogen lamp with calibration traceable to theNational Institute of Stan- dards and Technology (NIST). A stability test ,isperformed weekly using a 45-W lampin a light-tight hous•g thatmounts directly on the instrument on the concrete pad. A wavi:length calibration is also performed weekly using a mercury lamp, per- mitt•g a wavelength precision of:1:0.1 nm. The wavelength re- pea•ability is excellent (:•-0.02 nm) because each scan is aligned against the known Fraunhofer spectrum using a correlation alignment method. The expected long-term accuracy of the spectroradiometer system is expected to be better than a:5%. A detailed error analysis forthis instrument was given by McKen- zie et al. [1992]. 2121