We review the observations and spectral analysis for daytime measurements of the vertical column abundance of hydroxyl (OH) in the earth's atmosphere from the Fritz Peak Observatory in Colorado (40°N) and from a concurrent series of observations from New Mexico Institute of Mining and Technology (34°N). These are high-resolution measurements of solar ultraviolet absorption by atmospheric OH in the P1(1)2Π→2Σ electronic transition at 3081.7Å. The Fritz Peak OH database, initiated in 1977, consists of over 19,000 measurements and shows distinct diurnal, seasonal, and long-term variations. New Mexico OH observations began in 1996 using an instrument of comparable design and operation. Data from both locations are in conflict with OH abundances calculated by photochemical models for solar zenith angles less than about 60°. In addition, OH variations observed with respect to latitude, season, and long-term changes are not explained using current models. We present a critical examination of instrument characteristics and spectral analysis, one which indicates no tendency for systematic or interference effects that could contaminate observed OH abundances or their variations. This includes results of sensitivity analyses on synthetic spectra, taking into account temperature and pressure variations along the atmospheric optical path, and effects of absorption by SO2 and CH2O.

A new regime of reduced hydroxyl (OH) column abundances at Fritz Peak Observatory, first observed in the fall of 1991, continues to develop. Beginning in 1991, morning OH abundances in the fall season at Fritz Peak (40°N, 105°W) were found to be about 15% lower than the previous 1980–1990 average. Measurements for July–October 1997 indicate a continued regime of depressed (25–30%) OH abundances. The August‐October afternoon abundances were also low in 1997, yielding a PM‐AM asymmetry of about 6% which is near the annual average for Fritz Peak. This is the first occasion since 1980 that the fall PM‐AM difference of about 35% has not appeared. The winter season maximum of 15%, a persistent feature of the 1980–1990 measurement record, also failed to appear in 1996–97.

The 18‐year data base for vertical column abundance measurements of atmospheric OH at Fritz Peak Observatory, Colorado (40°N, 105°W), exhibits a relatively abrupt change in seasonal behavior during 1991 through 1995. Late summer through early fall abundances, especially, for each year of the 1991–1995 period are significantly below the comparable seasonal average for 1980–1990. Early morning abundances for August through October at solar zenith angles greater than 45° exibit an average reduction of about 0.8 × 1013/cm², which at a solar zenith angle of 60° is a 15% column decrease.

Examination of odd-hydrogen production and loss processes in the lower stratosphere reveals that the large abundances of midlatitude ClO recently measured requires a reassessment of the OH yield from methane oxidation due to a more rapid initiation of the methane oxidation sequence by atomic chlorine. Using a steady state calculation for OH and an iterative procedure for evaluating odd-hydrogen production from the methane oxidation sequence, we demonstrate that both OH and reactive chlorine are amplified below 25 km. The amplified values of OH and ClO are found to be consistent with recent measurements of ClO concentrations and of OH column abundances.

The first Southern Hemisphere measurements of the vertical column abundance of atmospheric hydroxyl (OH) have been obtained at Lauder, New Zealand (45°S) with a PEPSIOS instrument measuring the absorption of sunlight at 308 nm. The variation of column OH with solar zenith angle is similar to that measured at other sites. However average annual abundances of OH are about 20% higher than those found by similar measurements at 40°N. Minimum OH abundances, about 10% less than average levels at 40°N, are observed during austral spring. The OH abundance abruptly increases by 30% in early summer and remains at the elevated level until late the following winter.

Measurements of the vertical column abundance of atmospheric hydroxyl (OH) have been made during the period 1987–1989 at the National Weather Service (NWS) station at Moen, Truk, Federated States of Micronesia (7°N, 152°E). A total of 384 independent data sets was obtained. Tropical OH abundance levels average about 22% above corresponding mid‐latitude values, with OH levels during late winter and early spring up to 50% above those observed at 40°N. Stratospheric wind and temperature data obtained from the daily NWS radiosonde data are examined for correlations with the OH results.

OH vertical column abundances measured at Fritz Peak Observatory, Colorado, from 1977 through 1988 reveal semi‐annual and annual cycles which are amplitude modulated over the 11‐year data base. The modulation of the OH seasonal behavior is in phase with solar activity as described by the sunspot number index. No mechanism for the observed covariance of this OH behavior with solar activity has yet been identified.

The preliminary results of ground-based OH column abundance measurements from Truk, Federated States of Micronesia, are contained. These are the first OH column measurements from the tropics, and constitute a signficcant contribution to the OH data base. Comparisons of tropical OH behavior with the extensive mid-latitude observations serve as a critical test of the current understanding of the HO (sub x) photochemistry and its relationship to the other major chemical families. The quasi-biennial oscillation (QBO) in tropical stratospheric winds exerts a major influence on the Hadley cell vertical transport. Related QBOs in total O3 and in stratospheric H2O were identified, but QBO effects on other stratospheric species are still unknown. The solar tide in the tropics produces a diurnal surface pressure variation of 2 to 3 mb; its effect on OH photochemistry in the stratosphere may be significant.

Hydroxyl (OH) vertical column abundances measured with Pepsios spectrometers from Fritz Peak, Colorado (40°N), Boca Raton, Florida (26.4°N), and Poker Flat, Alaska (65°N), are presented. The Colorado normalized abundances have remained nearly constant since 1980 at levels about 1.7 times the 1977 values. Theoretical models agree well with the pre‐1980 abundances, but recent data appear to require a significant change in the atmospheric photochemistry. The Colorado seasonal diurnal asymmetry and its correlation with total ozone are shown to persist. The Florida data show little seasonal variation and have diurnal asymmetry characteristics which differ considerably from those of the Colorado asymmetry. Large Florida OH abundance excursions with respect to Colorado levels are seen in the wintertime of 1980, 1984, and 1986, which suggests the possibility of a relationship with the quasi‐biennial oscillation in tropical stratospheric winds. Responses of the Florida OH column to the passage of a tropical storm (July 1985) and to a partial solar eclipse (October 1986) are also documented. An anomalous drop of 80% in OH column abundances on March 18, 1986, is discussed. Hydroxyl abundance values at Poker Flat (June–July 1983) are 40% higher than corresponding Colorado or Florida values.

A large amplitude oscillation in the vertical column abundance of atmospheric OH has been observed in ground-based spectroscopic absorption measurements from Fritz Peak Observatory, Colorado (105°W, 40°N) during and after the partial solar eclipse of May 30, 1984. An initial OH reduction during the eclipse was followed by an underdamped oscillation having a period of about one hour; the OH abundances returned to normal values two hours after the eclipse termination. This is believed to be the first observation of a "ringing" response of any atmospheric constituent to a solar eclipse.

The OH vertical column abundances measured with the Pepsios spectrometer from the National Oceanic and Atmospheric Administration Aeronomy Laboratory's Fritz Peak Observatory (40°N, 105°W) for the 1977–1982 period are presented. Analysis of the ground state absorption shows a general seasonal variation of 28% with maximum in January and minimum in July. An A.M.‐P.M. asymmetry in the observations is found which has a 27% seasonal reversal, which is in phase with total ozone. A unique perturbation of +30% is found for summer 1982 OH abundances, which is attributed to the El Chichon eruption in April 1982.

An ultraviolet spectrometer of Pepsios design has been constructed and used for measurements of the vertical column abundance of atmospheric hydroxyl. Ground-based observations are made of the spectroscopic absorption of sunlight by OH at 3081.7 Å. The measurements are of relevance to the problem of stratospheric ozone. The spectrometer is a series arrangement of four pressure-scanned Fabry-Perot etalons with vernier spacer ratios. The spectral resolution of 0.06 cm−1 permits the identification and measurements of the sharp absorption feature from cool terrestrial OH against the solar background. The observations from Fritz Peak Observatory, Colo., are presently contributing to a new data base on this atmospheric trace constituent which should be of importance in the understanding of middle atmospheric photochemical processes.

New measurements of the vertical column abundance of atmospheric hydroxyl have been taken at Fritz Peak Observatory, Colorado, through the period of maximum solar activity of Solar Cycle 21. The OH abundance increase of about 12%/year observed since 1976 ended with the peak solar activity in 1980. The overhead‐sun column abundance increased from 5.4 × 1013/cm² in 1977 to 9.5 × 1013/cm² in 1980. The 1981 observations show a subsequent average decrease of about 3%. The seasonal average abundances for the entire 1976–1981 period show a good linear correlation with the corresponding sunspot number averages. A high‐sun enhancement of about 2 × 1013/cm² occurred during the 1979–1980 period; the resulting steep sec χ‐distribution has persisted into 1981.

Extensive new information on the daytime column abundance of atmospheric OH is presented from Spectroscopic observations of sunlight absorption by the P1(I)A2Σ+ – X2π(0, 0) transition of OH at 3081.7 Å, taken with a Pepsios spectrometer. The observations were made from the NOAA Aeronomy Observatory at Fritz Peak, Colorado, between December 1976 and December 1979. An overall seasonal dependence on the depth of penetration of the solar ultraviolet flux, which is believed to initiate the photochemistry of daytime OH, was found. The overall averaged data is represented by an empirical curve, N(sec χ), where χ is the solar zenith angle, showing an overhead sun maximum abundance of 7.1 × 10l3/cm2, decreasing to 4.9 × 1013/cm2 at sec χ = 2.0. The OH abundance and the sec χ variation agree reasonably well with predictions given by recent theoretical models of the stratosphere and mesophere and with the analysis of earlier in situ measurements of stratospheric and mesospheric OH by Anderson. An OH abundance increase of about 1 × 1013/cm2 per year is found for the 1976–1979 period. There was a seasonal variation in 1978 with a 25–30% decrease from a springtime maximum to a fall minimum. The 1978 abundances also show a 30–40% diurnal oscillation whose cause is as yet undetermined. This effect is shown to be dependent on solar flux on both a diurnal and an annual basis, and the 1978–1979 observations exhibit additional long‐term changes, possibly related to the solar cycle, which will be monitored in future work.