The SIEVERT system: taking into account GCR and SPE effects aboard aircraft

N. Fuller a, P. Lantos a and J.F. Bottollier-Depois b

Recently, the European Commission (EU directive 96/29/EURATOM) included the exposure of aircraft crew to cosmic radiation as occupational exposure. It was following the recommendations of the International Commission on Radiological Protection (ICRP, 1991) concerning the exposure to enhanced or elevated levels of radiation from natural sources. The effective dose should not be higher than 100 mSv over 5 years with a maximum of 50 mSv for a given year (specific rules apply to pregnant air crew).

The radiation doses onboard aircraft are due to two sources: Galactic Cosmic Rays (GCR) and Solar Proton Events (SPE). The doses are the result of the numerous secondary particles created in the atmosphere by high energy primary particles. The galactic component is permanent but modulated by the solar activity in the course of the 11-year solar cycle. The modulation parameter is an input of models such as EPCARD (Schraube, 1999) which computes the dose for GCR at any point in space up to an altitude of 80,000 feet. The SPE, when detected at ground level by neutron monitors (GLE), may enhance significantly the doses received onboard aircraft. A specific semi-empirical model named SiGLE was developed (Lantos & Fuller, 2003) to take into account these events.

Using EPCARD and SiGLE, the computerized system for flight assessment of exposure to cosmic radiation in air transport, or “SIEVERT” (Bottollier-Depois, 2003), is proposed to airline companies for assisting them in the application of this new legal requirement. This dose assessment tool was developed by the French General Directorate of Civil Aviation (DGAC) and partners: the Institute for Radiation Protection and Nuclear Safety (IRSN) and Paris Observatory. This professional service is accessible to airlines but also to a larger public via the internet site www.sievert-system.org, which allows any passenger to get an estimate of the dose received during a given flight.

The IRSN updates the map of the dose rates every month by taking solar activity into account. A map of the hourly dose at a typical subsonic altitude is given as an example for January 2005. In the case of a GLE, a specific map is created (see below). In addition, regular radiation measurements, from dosimeters installed on the ground and on aircraft, are used to confirm and, if necessary, to correct the obtained values.

The company prepares a file of completed or scheduled flights, and sends it to the SIEVERT Internet address. The system then completes the file by adding the effective dose that corresponds to each flight. Doses are calculated according to flight characteristics, using the dosimetric input data validated by the IRSN. It is asked to airlines to described a flight using way points. If the information is minimal (like information available on flight ticket), the dose value is assessed using a standard route profile. The data, at this stage, are anonymous. Airlines are in charge to add up the doses received during flights carried out by each member of the flight personnel.

The SIEVERT principle

Airspace is divided into cells. Each one is 1000 feet in altitude, 10° in longitude and 2° in latitude. Altogether they form a map of 265,000 cells; to each of these cells is assigned an effective dose rate value. The time spent by the plane on each cell and the corresponding dose are calculated; their accumulative total gives the dose received during the flight.

Validation

Results show that a monthly cartography based on the average intensity measured with a neutron monitor is sufficient to achieve a precision of about 20 % on effective dose calculation, for each flight. This study also pointed the importance of using the detailed flight plan of each flight to achieve sufficient precision. Indeed, on a subsonic route like Paris-Washington, two flights, operated on the same month, with the same aircraft, on the same route and direction could exhibit a relative variation of more than 50 %.

Application to past and last GLE

The bar plot below summarises the effective doses received for two routes during 31 GLEs (over 67 observed until 2004), the others giving negligible radiation effect. To each GLE correspond four bars. The first (in black) is the contribution to effective dose of the GLE itself for Paris-New York flight on-board Concorde. The second (in white) is the total effective dose taking into account GCR contribution too, calculated for the month of the event. The two last bars are the same but for Paris-San Francisco subsonic flight. All calculations correspond to the worse case in terms of departure time. It should be noted that the lower protection at supersonic altitude is counterbalanced by the flight durations which are quite different: 11 h 24 m for subsonic flights instead of 3½ h for Concorde. This explains the rather small difference observed between black bars for a given GLE. According to these results, over the 67 GLEs observed since 1942, only 18 are to be included in operational dose calculations, if we consider that the GLEs below 30 µSv could be neglected (this limit is representative of the lower limit of the effective dose received from GCR during a typical intercontinental journey).

The GLE 68 of the 20 January 2005 showed a very important North-South anisotropy above 65 ° in geomagnetic latitude. It was measured at an intensity of 178.4 % with Kerguelen NM, 3308 % with Terre Adélie NM and 2091 % with McMurdo NM (with 5 minute counts). In the North hemisphere, at about the same geomagnetic latitudes, the intensity is only 277% for Inuvik (Canada) NM, 114 % for Thule (Greenland) NM and 112 % for Barentsburg (Spitzberg) NM. It thus appears as one of the strongest GLEs observed during the last fifty years.The following table gives doses received from GLE 68 and from galactic cosmic rays for a few typical flights. The flights are based on actual flight plans and doses are calculated with the SiGLE model. The doses received from galactic cosmic rays (GCR) are calculated with CARI 6 software.

Doses obtained from measurements by IRSN between 1996 and 1998. The circles contain the average dose equivalent rate on the flight in µSv/h, and the radiation mean quality factor. The total dose equivalent (mean rate x time) is given for a round-trip flight.

History of significant GLEs in term of dose since 1942 for a supersonic and a subsonic flight (see text for full description)

Principles of data exchange between SIEVERT and airline companies.

Dose calculation principle with SIEVERT

Conclusion

SIEVERT provides a correct application of the regulation for at least three reasons. First, the results obtained are close enough to reality to avoid under-estimating the doses received by the personnel. Second, the radiation dose assessment mode is the same for all airlines. Third, if checks become required in the future, retrospective dose calculations might always be performed. A pioneering aspect of SIEVERT lies in the fact that it takes both potential radiation sources into account, GCR and SPE, using two efficient tools, EPCARD and SiGLE, which have been tested and validated. The system is used in routine at a national level since 2000. About 70,000 flights per month are proceeded by the overall French airlines.

The SiGLE principle

The semi-empirical model SiGLE combines few available measurements obtained on board Concorde during GLEs in 1989 and 2000 and on board a subsonic flight during a GLE in 2001, with calculations based on particle transport codes for GLE 42 on 29 September 1989, to compute an estimate of the dose D(t) received during GLEs .

D(t) = A(z, ) x L(G) x C( ) x I(t)

From the Air France and British Airways Concorde measurements, a linear relationship C() between ground based neutron monitor GLE time profiles and dose rates at 60000 in altitude is derived for different particle rigidity spectral exponents (noted γ). The rigidity spectrum exponent is deduced from the ratio between two neutron monitors (Webber &Quenby, 1959 ; Lantos, 2005) when an complete calculation is not available (recent GLEs for example).

The measurement on board a Czech Airlines flight from Prague-to New York (Spurný & Dachev, 2001), during the GLE numbered 60, on 15 April 2001, as well as plots based on theoretical calculations by O’Brien et al. (1998), are used to derive the attenuation factor A(z,) between dose rate at 60000 feet in altitude and dose rate at the aeroplane altitude, noted z. Because the available flights of Concorde were restricted to routes between New York (geomagnetic latitude λG = 50.7°N) and Paris (λG = 51.1°N) or London (λG = 53.7°N), the

computed doses are valid for the North Atlantic path.

L(λG) function, giving the variation of the dose rate

with the geomagnetic latitude at subsonic altitudes, is estimated using results of dose rate calculation during GLE 42 (O’Brien & Sauer, 2000) for Greenwich meridian. Then, from the results on North Atlantic path, the dose rates are deduced for other geomagnetic latitudes.The reference monitor for the model is Kerguelen Islands, in South Indian Ocean. It is located at Port-aux-Français (λG = 57.5°S and vertical cut-off

rigidity of 1.1 GV) . It is operated by the French Institute for Polar Research (IPEV)

Logarithm of the attenuation of dose equivalent rate in function of altitude for different values of the rigidity spectrum exponent . Attenuation for galactic cosmic rays is indicated with dashed line and attenuation for GLEs with average rigidity spectrum exponent = - 4.7 is indicated with a dotted line.

Dose equivalent rate coefficient L in function of the geomagnetic latitude for subsonic altitude 35 000 feet. The reference latitude corresponds to North Atlantic routes. Lower axis gives corresponding vertical cut-off rigidity for northern hemisphere and European sector (epoch 1995).

The flights from Paris to San Francisco and the flight from Tokyo to Paris along polar route are specifically corrected from the anisotropy mentioned above (SiGLE gives respectively 96.9 µSv and 88.3 µSv without correction).The world map gives an example, at subsonic altitude, of the hourly dose computed with SiGLE at the time of the maximum of the GLE (GLE + GCR).

a Paris Observatory, 92190 Meudon, FRANCE

b IRSN, B.P. 17, 92262 Fontenay-aux-Roses, FRANCE