News & Events

Red sky at night, shepherd's delight; red sky at lunchtime?...

As the old saying goes, seeing a red sky at sunset is supposed to herald fair weather on the following day. On the 16th October 2017 the saying needed to include a new clause as the ex-Hurricane Ophelia reached the shores of the British Isles bringing exceptionally strong winds and rainfall to much of the Western UK, but also a peculiar haze to the skies of Central and Eastern UK with reports of the sun turning red for a time.

Hurricane Ophelia had already made the news as being the most-Easterly hurricane ever recorded driven by the above-average temperatures in the waters of the North-Eastern Atlantic Ocean. As the system moved north over the cooler waters nearer to the UK the system was downgraded to a post-tropical cyclone, but maintained very high-winds across much of Ireland and the Western half of the UK. In southern Ireland winds speeds of over 190 km/h were reported at Fastnet Rock off Cork coast. During the morning and afternoon of the 16th October, the rest of the UK started to notice that the sky started to turn a yellow and orange. What caused this to happen? Two factors: scattering of sunlight by wildfire smoke and Saharan dust.

Wildfires affect many areas of the world and as well as the local hazards such as loss of life and damage to property, the largest fires can send aerosol and trace gases high into the atmosphere which can then readily be transported across the globe in a matter of days to weeks. In some cases smoke from intense fires can be injected directly into the stratosphere through pyrocumulus clouds and the consequences of these smoke particles on the radiation budget and stratospheric chemistry are still poorly understood.

Figure 1. Evolution of Hurricane Ophelia to an intense post-tropical cyclone, seen by Aqua MODIS true colour image on 15/10/2017. Red dots indicate fire hotspot detection.

Figure 1. Evolution of Hurricane Ophelia to an intense post-tropical cyclone, seen by Aqua MODIS true colour image on 15/10/2017. Red dots indicate fire hotspot detection.

This year there have been a number of severe wildfires affecting Portugal and Spain. During 15th October a number of large fires were reported across the region at a time when the ex-Hurricane Ophelia was transforming into a mid-latitude weather system. The MODIS instrument on Aqua (figure 1) showed this happening in the early afternoon of October 15th as the system passed close to the Portuguese coast and air from the region became drawn into the developing system. As the system tracked northwards towards Southern Ireland, the trailing cold front within the weather system acted as a conveyor belt to feed the combination of smoke and other pollutants directly from the fire region up towards the UK.

Hurricane Ophelia began its life-cycle on October 6th as a developing circulation to the South west of the Azores. Over the next few days the system remained fairly stationary, developing further over the warm ocean, into a category 2 hurricane on October 12th and by the 14th had been classified as a category 3 storm. During this period of development, a major dust outflow event from the Sahara desert was occurring. These sporadic dust events provide a major source of nutrients to the subtropical Atlantic Ocean but, in this case, a lot of the dust become entangled into the developing hurricane.

Figure 2. Synoptic weather chart with a false-colour infrared SEVIRI channel 9 image overlain at 6UTC on 16/10/2017.

Figure 2. Synoptic weather chart with a false-colour infrared SEVIRI channel 9 image overlain at 6UTC on 16/10/2017.

By the morning of 16th October at 6 UTC, the system, now classified as a post-tropical cyclone was beginning to bring strong winds to the south of Ireland and a weakening cold front was moving into the south-west of the UK with very little rain associated with it. The system moved in a predominantly northerly direction over the next 6 hours or so with the cold front moving very slowly eastward. It was during this time that many people across the UK were starting to report the sky turning yellow and the sun starting to look red. How is this explained and what could we see from Earth Observation satellites and ground-based observations?

The atmosphere consists of a number of gases, aerosols and particles which all interact to a stronger or lesser degree with incoming solar radiation and outgoing longwave radiation emitted from the Earth’s surface. For incoming radiation, scattering of the sunlight happens as it passes through the air and interacts with smoke or dust particles. The general effect is to diffuse the light, by spreading it out in all directions rather than just a single straight beam which would occur if there were no atmosphere. There are three different types of scattering: Rayleigh scattering, Mie scattering and non-selective scattering. The Mie scattering effect explains why clouds are white, for example. Clouds contains billions of small water droplets (which are themselves clear) which form around a nucleus which could be dust, smoke or another particle. As sunlight interacts with the newly formed droplet, the light is scattered. As the cloud contains so many droplets the light is scattered many times, an effect called multiple scattering, and this causes the colours of the light to recombine to make white light to an observer on the ground. So why was the sky yellow rather than white?

From ground-based observations with the Chilbolton lidar on 16th October (figure 3) the haze layer was actually made up of several very thin layers between 4 and 7 km in altitude. As the smoke and dust particle layers were very thin, multiple scattering would likely be less important and single scattering the dominant mode (i.e. only one, or at least very few, interactions between the visible light and the particles). Blue wavelengths are scattered more than the red ones and this means that many of the blue and green wavelengths will be scattered directly back into space with only the yellow and orange wavelengths being observed at the ground.

Figure 3. Lidar observations at Chilbolton station on 16/10/2017. The observations show thin layers of increased backscatter between 4 and 7 km

Figure 3. Lidar observations at Chilbolton station on 16/10/2017. The observations show thin layers of increased backscatter between 4 and 7 km.

MODIS satellite image from 5km taken 16/10/2017

From Earth-observation satellites, the VIIRS visible channel satellite data at around 13:30 UTC on the 16th October (figure 4), clearly showed the smoke/dust layer as a brown feature (clouds showing as white in the same figure). The plume extended from the Bay of Biscay up to Scotland with a width of up to 250 km, which explains why many parts of the UK experienced the same phenomenon. By this time, it can also be seen that the system was occluding rapidly and so the winds, although still strong, were beginning to abate across Ireland and Scotland.

Figure 4. VIIRS visible channel image of North-West Europe at ~13:30 UTC on 16/10/2017 when many people in the UK were reporting yellow skies and a red sun.

Figure 4. VIIRS visible channel image of North-West Europe at ~13:30 UTC on 16/10/2017 when many people in the UK were reporting yellow skies and a red sun.

Figure 5. Copernicus Atmospheric Monitoring Service model forecasts of (left) dust aerosol optical depth and (right) biomass burning optical depth at 12 UTC on 16/10/2017. Rough estimates show a 70% biomass burning aerosol and 30% dust aerosol.

Figure 5. Copernicus Atmospheric Monitoring Service model forecasts of (left) dust aerosol optical depth and (right) biomass burning optical depth at 12 UTC on 16/10/2017. Rough estimates show a 70% biomass burning aerosol and 30% dust aerosol.

Estimating the separation between the different aerosol types involved in the Ophelia system is a complex task. Fortunately model predictions existed from the Copernicus Atmospheric monitoring service which had done a very good job of predicting the system track towards North-West Europe and the separation of dust, smoke (and sea salt in the core of the system). Using optical depth predictions as a guide, the yellow sky across the UK was caused by a mixture of 70% biomass burning smoke and 30% Saharan desert dust. These estimates will likely be refined once satellite observations are considered.

Figure 6. Aerosol absorbing index values from the Metop-B GOME-2 instrument on the morning of 16/10/2017.

Figure 6. Aerosol absorbing index values from the Metop-B GOME-2 instrument on the morning of 16/10/2017.

Satellite observations will play a key role in the post-analysis of the event, and early analysis of the GOME-2 on Metop-B aerosol absorbing index (figure 6) match the model predictions of where the highest aerosol was likely to occur across much of the UK. Observations of another pollutant, carbon monoxide (figure 7), show some very high values (over 140 ppb) across much of Central and Southern England. This gas is itself an indirect greenhouse gas and has an important effect in regulating levels of methane in the atmosphere as both share a common removal process (reaction with the OH radical). Again, more analysis will be needed to determine at which altitude the highest aerosol and carbon monoxide was occurring, but based on the lidar observations it was likely to be between 4 and 7 km and so should have had little effect on air quality near to the surface.

Ex-hurricane Ophelia was quite a remarkable system over North-west Europe, being the most Easterly hurricane on record and only forming originally due to a combination of favourable atmospheric conditions over the Atlantic Ocean and the fact the water was up to 2 degrees Celsius warmer than average. Ireland experienced some of the strongest wind gusts in the country’s history whereas much of the rest of the UK reported yellow skies and a red sun due to the system interacting with both uplifted Saharan dust and Portuguese and Spanish wildfires. It is only by the whole suite of satellite data currently available that these systems can be monitored and analysed and we gain a new understanding of the world around us.

Figure 7.  Aqua AIRS/AMSU measurements of carbon monoxide on the afternoon of 16/10/2017 with values over 140 ppb recorded over central and southern England.

Figure 7. Aqua AIRS/AMSU measurements of carbon monoxide on the afternoon of 16/10/2017 with values over 140 ppb recorded over central and southern England.

Dr David Moore is a Researcher at the National Centre for Earth Observation, University of Leicester. Twitter: @David_P_Moore