Why add extra minutes to Sunset?

Why do we need to add 5 minutes to sunset times? 

(Mufti) Amjad Mohammed

(Maulana) Yousaf Baig

(Maulana) Farid Patel

I have seen some Ula­ma rec­om­mend­ing to add 2–4 min­utes to sun­set times and oth­er Ula­ma strong­ly con­demn­ing this rec­om­men­da­tion? Why should 2–4 min­utes need to be added to absolute times obtained from Roy­al Green­wich Obser­va­to­ry?

In the Name of Allah, the Most Gracious, the Most Merciful.

As-salā­mu ‘alaykum wa-rah­mat­ul­lāhi wa-barakātuh

We would like to begin by extend­ing our grat­i­tude to you for ask­ing this ques­tion. May Allah (SWT) give you the best of rewards (Ameen).

We are assum­ing that you are using accu­rate fig­ures from a rep­utable site such as HMNAO. We rec­om­mend that you add 5 min­utes to the list­ed sun­set times, for the rea­sons explained below.

[2:187] And eat and drink until the white thread of dawn becomes dis­tinct to you from the black thread [of night]. Then com­plete the fast until the sunset

Refraction of Light:

When light pass­es from one medi­um (mate­r­i­al) to anoth­er, it changes speed. This is because the speed of a light wave is deter­mined by the medi­um through which it is pass­ing. This change in speed caus­es the light to bend (refract) and it is a well known and observed phe­nom­e­non. Observ­able refrac­tion is demon­strat­ed in the fol­low­ing picture.

Light from the sun is sim­i­lar­ly impact­ed (bent) on its way to earth by sev­er­al fac­tors. Pol­lu­tion with­in our atmos­phere affects the refrac­tion of light 1 . and is dif­fer­ent at var­i­ous places on Earth. In addi­tion to pol­lu­tion, the extent of the refrac­tion also depends on atmos­pher­ic tur­bu­lence, includ­ing air tem­per­a­ture and atmos­pher­ic pres­sure: the high­er the pres­sure and the low­er the tem­per­a­ture, the larg­er the refrac­tion angle. So, if you watch the sun set in an area of high pres­sure on a cold day, you may have to wait sev­er­al sec­onds for the upper edge of the sun to dis­ap­pear behind the hori­zon, com­pared to a day with aver­age pres­sure and tem­per­a­ture 2. Although pol­lu­tion may have an impact on refrac­tion, tur­bu­lence in the Earth’s atmos­phere has a much high­er impact on refrac­tion of light, par­tic­u­lar­ly tem­per­a­ture vari­a­tions in the atmos­phere. 3

When it comes to sun­set for the estab­lish­ment of Maghrib time, you have to account for refrac­tion because the phys­i­cal sun is actu­al­ly low­er than what appears to our eyes. 4

Why was 4 minutes recommended before?

The phe­nom­e­non of sunrise/sunset times not being accu­rate was first raised in the 1990s. In a research paper by Dr Bradley E. Schae­fer (1990) 5 he explic­it­ly states that “the time of sun­rise can only be pre­dict­ed with an accu­ra­cy of 4 [min­utes]”, and sun­set times may also suf­fer from inac­cu­ra­cies. Sci­ence is dri­ven by obser­va­tions and research and in 2018, we have more pre­cise obser­va­tion­al data for high lat­i­tude countries.

(Standard) Refraction

We have noticed that in some of the pro­grams stan­dard refrac­tion is already account­ed for e.g. HMNAO data sheet (7) states:

The sun­rise and sun­set times that are pre­pared by Her Majesty’s Nau­ti­cal Almanac Office cor­re­spond to the instants times when the zenith dis­tance of the cen­tre of the Sun is 90° 50 ́: this allows 34 ́ for refrac­tion in the atmos­phere and 16 ́ for the angu­lar semi-diam­e­ter of the Sun. (The zenith dis­tance of an object is the angle between the direc­tion to the zenith (i.e. the point ver­ti­cal­ly over­head) and the direc­tion to that object.) 6

Time and Date AS states the fol­low­ing about their account­ing of the refrac­tion:

Our Sun Cal­cu­la­tor, which shows sun­rise and sun­set times in a loca­tion of your choice, takes the effect of refrac­tion into account. Our cal­cu­la­tions are based on the stan­dard atmos­pher­ic pres­sure of 101.325 kilo­pas­cals and tem­per­a­ture of 15°C or 59°F, so sun­rise and sun­set may hap­pen some sec­onds before or after the stat­ed time when the tem­per­a­ture or pres­sure strong­ly devi­ate from the average.

Here are the sunrise/sunset times for a loca­tion in Brad­ford which is at 53°47’38.08“N, 1°45’7.42“W:

Date and Time05:4020:30
Islam­ic Finder05:4020:31

Variance in (calculated) sunset timings across geographies & seasons

While there are sci­en­tif­ic stud­ies on this top­ic from var­i­ous regions around the world, there is no con­sen­sus on how refrac­tion affects sun­set times in (spe­cif­ic parts of) Britain. 7. Hav­ing said that, raw data from Alber­ta, Cana­da has demon­strat­ed a sea­son­al change in the obser­va­tion of sun­set with greater lev­els of refrac­tion around win­ter com­pared to sum­mer 8  When this refrac­tion data is used in the U.K. set­ting we find the val­ue vary­ing from 4 min­utes to 7 min­utes. We appre­ci­ate this may not be whol­ly accu­rate due to the dif­fer­ence in set­ting, but until a sim­i­lar study is con­duct­ed in U.K. we will utilise it; espe­cial­ly as it sug­gests a lat­er time which is more pre­cau­tion­ary when it comes to acts of wor­ship. For instance, we can add that sim­i­lar research car­ried out in Hole­town, Bar­ba­dos 9 found the refrac­tion to be low­er sug­gest­ing a fur­ther two min­utes deduct­ed from HMNAO would be suf­fi­cient. Return­ing to the Cana­di­an data, once the two min­utes which HMNAO deduct to take Astro­nom­i­cal refrac­tion into con­sid­er­a­tion we are left with an addi­tion­al 2 to 5 min­utes. We have con­sult­ed two experts in this field and their opin­ions have dif­fered on the mat­ter. There­fore to take sea­son­al effects into con­sid­er­a­tion and to be as pre­cau­tion­ary as pos­si­ble with­out increas­ing the time unnec­es­sar­i­ly, also to issue a stan­dard val­ue for the year we rec­om­mend adding five min­utes to the HMNAO time for sun­set. We have con­fi­dence from those astronomers that we have con­sult­ed that we will be well with­in the cor­rect time­frame. This is an ongo­ing research pur­suit for Wifaqul Ula­ma and more research will be pre­sent­ed in due course.


There­fore, we sug­gest that the addi­tion of five min­utes to the HMNAO time to be more pre­cau­tion­ary and hence rec­om­mend that based on cur­rent understanding.

جزاك اللهُ خيرًا

Dr Victor P. Debattista

Hi Dr. Debattista, 

I spoke to you a cou­ple of weeks ago on the phone. We do have a tech­ni­cal query and we were won­der­ing if you can answer it for us.

We have looked at HMNAO sun­set times and under­stand that they account for refrac­tion in their cal­cu­la­tions. How­ev­er, we believe the cal­cu­la­tions to be stan­dard and static. 

The Islam­ic tra­di­tion is to actu­al­ly observe the sun­set but nor­mal­ly Mus­lims take their times from the Observatory. 

In clas­si­cal appli­ca­tion, a per­son would be observ­ing the sun­set and then open­ing their fast when the sun is no longer vis­i­ble. Mus­lim need to use the sun­set times to open their fasts dur­ing Ramad­han (after sun­set) and to pray etc. In your opin­ion, is our rec­om­men­da­tion of adding two min­utes suf­fi­cient to the HMNAO times to account for the impact of refrac­tion? This query is specif­i­cal­ly for the times for Britain.

Our rec­om­men­da­tion is pure­ly as a pre­cau­tion­ary mea­sure to ensure that the sun­set has occurred. 

Does air pol­lu­tion have any impact on HMNAO sun­set times? 

Can we pub­lish your answer on our web­site with your answer?

We have asked this of a num­ber of experts and will be look­ing at all of the replies comprehensively. 



Thank you for your question.

The cal­cu­la­tion that is being done by HMNAO, as per your link, requires that the Sun is half a degree below the hori­zon, which they deter­mine means that the Sun is no longer vis­i­ble even with refrac­tion tak­en into account. The thing to keep in mind about refrac­tion is that it is high­ly depen­dent on atmos­pher­ic con­di­tions. Not so much pol­lu­tion as tur­bu­lence (wind etc) and, espe­cial­ly, tem­per­a­ture vari­a­tions in the atmos­phere. These are basi­cal­ly weath­er pat­terns and, as usu­al, vary sig­nif­i­cant­ly by day and by loca­tion and are essen­tial­ly impos­si­ble to pre­dict with cer­tain­ty. A search online showed that refrac­tion as large as 2 degrees has been record­ed historically!

It is worth point­ing out that this was in Antarc­ti­ca where one might expect tem­per­a­ture vari­a­tions in the atmos­phere to be more extreme.

So my con­clu­sion as an astronomer is that allow­ing a fur­ther two min­utes for the Sun to set is a good idea as a pre­cau­tion but that one should always be aware that extreme (there­fore unlike­ly but not impos­si­ble) con­di­tions may result in the Sun reap­pear­ing when it should be well below the hori­zon. Anoth­er issue to be aware of is that sun­set varies with alti­tude. At 12 km above the sur­face, as on a plane cruis­ing, the hori­zon is 2 degrees low­er, which means the Sun sets about 1.5 min­utes lat­er (in the UK). This prob­a­bly affects a small num­ber of peo­ple but it might be worth draw­ing your atten­tion to it.

I hope this helps with your deci­sion mak­ing. Please do let me know if you need more help with this and I will try my best to help you.

best wish­es,


Yes, the cor­re­la­tion between sun­set and lat­i­tude and the extreme sce­nario of the sun reap­pear­ing again is noted.

How­ev­er we are mak­ing an assump­tion that the per­son is at ground lev­el in the British isles and observ­ing sun­set on a typ­i­cal day with no abnor­mal deviation.

Based on our assump­tions, we would like to rec­om­mend adding extra two min­utes to HMNAO times to British Mus­lims and I hope that in your opin­ion this will be sufficient?

And can we add your reponse with your name to our site?

Many Thanks


Indeed, adding a fur­ther two min­utes would ensure that any fluc­tu­a­tion due to tur­bu­lence, pol­lu­tion or oth­er fac­tors is com­pen­sat­ed for. You can add my name to the site. 

Best wish­es,

(Dr) Vic­tor P. Debattista
Pro­fes­sor of Astrophysics

Professor Andy Newsam


Thank you for get­ting in touch — that is a very good ques­tion! Unfor­tu­nate­ly, there is no sim­ple answer.

First of all, it depends a lit­tle bit on exact­ly what you mean by “sun­set”. The obvi­ous def­i­n­i­tion is sim­ply “when the sun has dis­ap­peared below the hori­zon”, but that will clear­ly depend on your sur­round­ings — sun­set will appear to be much ear­ly if you are in a val­ley sur­round­ed by moun­tains, than if you are look­ing out to sea. In astron­o­my, we tend to choose a more con­sis­tent def­i­n­i­tion which is, in effect, “when the sun drops below the hori­zon that you would have if you could see the sea”. That still depends on your alti­tude (it will be a bit lat­er if you are high­er up) but it is more con­sis­tent and can be cal­cu­lat­ed accu­rate­ly for any place on the sur­face of the Earth if the alti­tude is known.

How­ev­er, there are some things that you can­not cal­cu­late in advance. The main one is refrac­tion, as you say. When the sun seems to dis­ap­pear below the hori­zon is actu­al­ly lat­er than when it real­ly did drop low enough in the sky, as the light is bent by refrac­tion through the air around the sur­face of the Earth and so you can still see the Sun a bit after it should have been blocked by the hori­zon (for­give me if I am telling you things you already know — I just want to make sure we are both talk­ing about the same thing).

Unfor­tu­nate­ly, the amount of refrac­tion depends on a num­ber of fac­tors, some of which can­not be pre­dict­ed. For exam­ple, the air pres­sure is impor­tant since that changes the “amount” of air between you and the hori­zon and hence the amount of refrac­tion. There­fore, when cal­cu­lat­ing sun­set time, we have to make an assump­tion. The sim­plest thing to do is assume there is no air (and hence no refrac­tion) but that is very inac­cu­rate, so we nor­mal­ly just assume a “stan­dard” air pres­sure. That gives an answer which is cor­rect on aver­age, but will be a bit inac­cu­rate on any giv­en day.

To make things more com­pli­cat­ed, as you move away from the equa­tor, the angle that the Sun sets moves fur­ther away from the ver­ti­cal (com­pared to the hori­zon) and changes dur­ing the year, so the amount of dif­fer­ence in sun­set-time that a giv­en amount of refrac­tion makes varies from place to place and dur­ing the year.

For astronomers, we tend to deal with this by being pes­simistic about sun­set times. Since we are only real­ly inter­est­ed in know­ing when the sky will def­i­nite­ly be dark, we tend to take the worst case, which around the UK usu­al­ly means a max­i­mum of about 6 or 5 min­utes being added to sun­set time from the “aver­age refrac­tion” cal­cu­la­tions. How­ev­er, such long delays are very rare — you are more like­ly to see no more than a cou­ple of min­utes extra delay, so your sug­ges­tion of adding 2 min­utes would usu­al­ly be fine.

Since this is not entire­ly pre­dictable (even with quite a good weath­er fore­cast) you are prob­a­bly going to have to decide what uncer­tain­ty you are com­fort­able with. If you want to be close to, but just after sun­set most of the time, then adding 2 or 3 min­utes to the HMNAO sun­set times will be fine. If you want to be cer­tain that sun­set will be over, then you may want to add 5 or even 6 minutes.

I hope that helps — feel free to share this if you think it will be use­ful for others.
Prof Andy Newsam
Direc­tor of the Nation­al Schools’ Observatory

References & Acknowledgements