Contents

A Satellite Survey of Water Vapor and Cirrus Cloud Cover in Northern Chile by Dr. D. A. Erasmus
Wind speed and direction at observatories and previous sites in Northern Chile
Photometric Skies for Paranal,Tololo, Mauna kea and Mt. Graham

Links

CTIO Weather and Environmental Information
ESO La Silla Meteo Monitor
ESO Weather Statistics
ESO Paranal Astroclimatology
Las Campanas Weather Page
ALMA Chajnantor Weather summary 1998-1999
ALMA weather data at Chajnantor

Since the performance of large telescopes at optical and infra-red wavelengths is critically dependent on atmospheric cloud cover and water vapor, a quantitative survey of these conditions at candidate telescope sites is an essential precursor to site selection. In view of the CTIO/NOAO proposal to build a 25-30m telescope in Northern Chile, a knowledge of the climatology of water vapor and cloud cover in the area, would be of vital importance.

In this proposal a study is outlined to survey the temporal (seasonal and inter-annual) and spatial variations in water vapor and cirrus cloud cover in Northern Chile using a proven technique. The technique involves the use of weather satellite imagery taken at 6.7m. In studies carried out by the author for European Southern Observatories (ESO) (Erasmus and Peterson, 1996; Erasmus and Stanko, 1997; Erasmus and Maartens, 1999), a method was developed to derive meteorological parameters that relate to astronomical observing quality from these satellite data. The Upper Tropospheric Humidity (UTH) quantifies the amount of water vapor in the middle and upper troposphere (see section 3). In addition, the presence and thickness of high altitude (9-12km) cirrus cloud can also be determined. An important advantage of using satellite data is that it provides continuous coverage over a large area with good resolution thus enabling one to quantify microclimatological differences within the study area.

In this proposal a study of the area 22S to 32S and 66W to 72W from September 1991 to August 1998 is outlined. This area was selected to include the Andean Altiplano and additional coverage to the N, S, E and W. The period September 1991 to August 1998 was chosen since high quality 6.7m imagery from the Meteosat-3 (September 1991 - June 1995) and GOES-8 (July 1995 - August 1998) satellites are available for this period. The study will provide a climatology of water vapor and cirrus cloud cover over the area. Selected statistics of water vapor and cirrus cloud cover parameters will be computed for the area.

The main circulation features controlling the weather and climate of Northern Chile are shown in Figure 1. The diurnal cycle of heating and cooling exerts a strong control on conditions near the surface. From the point of view of astronomical site selection the primary effect is the lifting of the trade wind inversion during the day and lowering of the inversion at night. Moist air and clouds are trapped below the inversion so that locations on higher mountain peaks (above about 2000m) will remain cloud free and dry. A secondary factor related to the diurnal cycle is the initiation of convection (thunderstorms) near the continental divide. East of the divide the availability of low-level moisture and the absence of a strong capping inversion (such as the trade wind inversion west of the divide) favors convection. It is therefore reasonable to expect that as one approaches the divide from the west, cloud cover and precipitation from these storms will increase. This variation will be mapped in the proposed study.

There are also seasonal variations in the strength and position of the circulation features that determine moisture and cloud cover conditions in the region. The subtropical high pressure over the Pacific ocean influences the region throughout the year. Subsidence in the high creates a strong temperature inversion at about 1000m above the surface which traps low level stratocumulus clouds below it (notice the area of medium grey speckelled cloud in and around the high). The subtropical high is a semi-permanent circulation feature that only experiences minor position changes as the high develops east or west of its mean position. The high is responsible for the typically clear and dry conditions over Northern Chile.

In summer, due to southward movement of the high, the possibility of intrusions of moist air from the tropics increases. Moisture is injected into the middle and upper troposphere by convection in the Amazon basin and is circulated anticyclonically to the west and south. Northern Chile may be affected by this moisture and cloud under these conditions. The exact extent and frequency of occurrence of moisture and cloud from this source will be quantified in this study.

In winter, the pressure systems move northwards and the subtropics come under increasing influence from migratory wave-like systems that propagate from west to east in the prevailing flow. The high pressure area of the wave (the ridge) is warm and dry while the low pressure area (the trough) is cool and moist. Middle and upper tropospheric clouds can typically be found along the leading edge of the ridge and trough in association with surface warm and cold fronts. Trough and ridge development, which is usually slight over the Southeastern Pacific, may add a meridional component to the cloud cover and water vapor advection patterns. The survey proposed in this study will also quantify moisture and cloud from this source.

On an inter-annual basis, it has been established that there is a strong correlation between atmospheric moisture conditions in Northern Chile, including water vapor, cloudiness and precipitation, and the occurrence of El Niño - Southern Oscillation (ENSO) events. ENSO is a non-periodic inter-annual oscillation in the atmospheric and oceanic conditions that exist over the tropical Pacific Ocean. Under normal or typical conditions, the weather of the southeastern Pacific is dominated by the strong, semi-permanent south Pacific Anticyclone (high pressure) which gives rise to persistent southerly trade winds along the coast of South America that become easterly and extend westwards along the Equator to near the dateline (Figure 1). Under these conditions, strong upwelling is encouraged in the ocean off the coast of Chile, producing very cold sea surface temperatures. The cold water is dragged by the trade winds along its trajectory into the tropics. This cold water reduces evaporation rates and also stabilizes the lowest layers of the atmosphere. Thus the two main ingredients for the formation of cloud and precipitation - moisture and vertical motion - are absent in this area. Consequently, in the subtropics of Northern Chile and the adjacent ocean, one finds desert conditions.

The cold phase of ENSO known as La Niña or El Viejo is simply an intensification of these "normal" conditions in which the south Pacific high, the trade winds and upwelling strengthen thus producing colder than normal sea surface temperatures. Under these conditions, the weather in Northern Chile remains dry and clear.

Every two to seven years, with irregular periodicity, an anomalous warming of the sea surface occurs. This warming is coincident with a decrease in the strength of the south Pacific high and consequently the trade winds and also the amount of upwelling. This anomaly, which usually lasts from 12 to 18 months has become known as El Niño (the warm phase of ENSO) and directly affects the whole of the tropical and equatorial south Pacific Ocean. The anomalous ocean warming off the coast of South America promotes evaporation and destabilizes the lower layers of the atmosphere thus encouraging cloud formation and precipitation. Therefore, even though Northern Chile is typically dry and clear, there will be periods associated with the occurrence of El Niño, when increases in atmospheric moisture and cloud cover may be expected.

Clear evidence for a link between ENSO conditions and cloud cover in Northern Chile comes from a recent study by Sarazin (1997) of the cloud cover anomaly at La Silla and Paranal over the last 14 years. Cloud cover anomalies, which may be as large as 35% (15%) above or below normal at La Silla (Paranal), are negative under El Niño conditions and shift to become positive under La Niña conditions. Another important result from this study is that the strength of the ENSO event, as indicated by the magnitude of the Southern Oscillation Index, is proportional to the magnitude of the cloud cover anomaly.

The period 1991- present has experienced a wide range of ENSO conditions, making it a good period to analyze in order to determine seasonal and inter-annual variations in water vapor and cirrus cloud over Chile. The table below provides a summary of ENSO events over the period.
 
 

PERIOD	ENSO CONDITIONS
Sept., 1991 - June 1992	Moderate to Strong El Niño
July 1992 - December 1992	Normal
January 1993 - November 1993	Weak El Niño
December 1993 - September 1994	Normal
October 1994 - March 1995	Weak El Niño
April 1995 - February 1997	La Niña
March 1997 - Present	Strong El Niño

 Figure 2. Southern Oscillation Index (Pressure anomaly at Tahiti minus pressure anomaly at Darwin, normalized) for the period 1979-1997.

Positive values indicate a La Niña and negative values an El Niño. Climate Prediction Center, NOAA, (http://www.elnino.noaa.gov/) Figure 2 shows the Southern Oscillation Index for the period 1979-Present and shows that the range of ENSO events that occurred during the period 1991-present are representative of those over the entire period (approximately 30 years). For this reason, it may be expected that, by analyzing satellite data for this period, a reliable assessment of the effects of ENSO events on water vapor and cirrus cloud cover over Northern Chile may be made.

This study will involve the use of measurements of water vapor and cirrus cloud derived from meteorological satellites. A great advantage of satellites is that they can provide continuous coverage of remote areas where conventional meteorological observations are sparse or absent. These satellites measure meteorological parameters by passive remote sensing at different wavelengths.

In studies carried out by the author (Erasmus and Peterson, 1996; Erasmus and Stanko, 1997; Erasmus and Maartens, 1999), software was developed that computes measurements of water vapor in the middle and upper troposphere from Meteosat-3 and GOES-8 satellite imagery taken at 6.7m.

Water vapor in the atmosphere is absorbent at most infra-red wavelengths. The absorptivity for a given wavelength determines the layer in the atmosphere in which out-going terrestrial radiation will be absorbed and re-emitted by resident water vapor. Figure 3 shows the weighting functions for different infra-red channels and indicates that observations at 6.7m are sensitive to water vapor emissions from the layer between 600mb and 300mb (There is little or no water vapor above 300mb since the air is too cold to carry any moisture). The monochromatic emittance from this layer depends on the amount of water vapor in the layer and temperature. Temperature can be accounted for by using an observed or representative temperature-height sounding so that emittance is then only a function of the amount of the emitting gas, in this case water vapor, in the layer.

Using Plank's law it is possible to define a temperature called the brightness temperature (T) which would be the equivalent blackbody temperature of an object with the measured monochromatic emittance (E) using the following equation:

c1
E(T) = 5 [exp{c2/T}-1]

where c1 and c2 are constants and is the observation wavelength. From the water vapor brightness temperature, humidity values may be derived using an equation provided by the agency operating the satellite such as the following one used to process Meteosat-3 water vapor imagery:

UTH(%) = exp(42.21613 - 0.15578Twv )

Cos(latitude)

UTH is the Upper Tropospheric Humidity in % and Twv is the water vapor brightness temperature. UTH can easily be converted to other moisture quantities such as Precipitable Water Vapor (PWV). In a study by Erasmus and Stanko, 1997, it was found that the mean absolute difference between the measurements of PWV made at Paranal using a dark sky emissivity meter looking upwards, the Antofagasta Radiosonde and the Satellite is less than 1mm. This is within the range of respective instruments error. Further, since different sampling (time and space) is used by the respective methods, the level of agreement is remarkable.

Cirrus clouds and their thickness may also be inferred from the 6.7m imagery. The presence of cirrus clouds, typically found at about the 300mb level, will produce very low brightness temperature values. IR radiation from water vapor below the 300mb level would be absorbed and re-emitted at colder temperatures by the clouds, the amount of transmission through the cloud being dependent on cloud thickness. For more information on atmospheric remote sensing techniques, please refer to Rao et. al., 1990, p.203-224, Wallace and Hobbs, 1977, p.279-315 and Menzel and Purdom, 1994.

In view of the factors noted in section 3, it is proposed that water vapor and cirrus cloud cover measurements be based on 6.7m satellite imagery of the study area. For the period 1 September, 1991 to 30 June, 1995 data from the Meteosat-3 satellite will be used and from 1 July, 1995 to August 31, 1998 from the GOES-8 satellite. The Cooperative Institute for Research in the Atmosphere (CIRA) at Colorado State University (CSU) operates a first-order ground receiving station for these data and has a quality controlled data archive. The Meteosat-3 data has a spatial resolution of 10km and for GOES-8 it is 4km. Area scans are available every three hours. The data will be sampled so that the diurnal, seasonal (annual) and inter-annual cycles are suitably resolved.

Representative statistics of water vapor and cirrus cloud cover parameters for the study area will be computed. Included will be the mean UTH (PWV) and frequency of occurrence of opaque and transparent cirrus cloud for each pixel. Based on these statistics the 10 best locations (pixels) will be ranked in terms of each of the water vapor and cloud cover parameters. For these 10 best locations, further statistics indicative of the seasonal and inter-annual variations will be computed.

A final written report on the results of this analysis as well as significant findings from the study will be prepared and submitted. In addition, an oral presentation by the author will be scheduled at a suitable location.

The steps envisaged in this project are outlined below. The study will be carried out under the direction of the proposal author. Close co-operation will be maintained with CTIO/NOAO representatives throughout the proposed study.

Step Task Date

A. Data acquisition and code adaptation 7/1/99 - 12/31/99

Obtain satellite data and establish database

Adapt code to process satellite data files

B. Data analysis and report preparation 1/1/00 - 6/30/00

Compute and map water vapor and cirrus cloud cover

parameter statistics for the study area

Prepare a written report on the results and document

significant meteorological findings from the analysis performed
 
 

The wind in the North of Chile as can be seen by the wind roses below is mostly from the North and West. On the Chajnantor Plain the wind is predominately from the west most of the year round, this is probably due to the fact that up high over Chile the wind is westerly and the height of the chajnantor plain puts it in this flow. Since the general wind direction in Chile is North and West, it was decided to favor mountains that were clear of obstructions in these directions, since there would be less of a possibility of having mountain waves and turbulent air flowing over a potential site. Since wind speed is important, mountains that are close together normally create venturi effects between close peaks that can double or triple normal wind velocity. The following windroses show the prevailing winds at previously studied sites.