How does API calculate the dose rates?

Jan Helebrant

unread,

Nov 10, 2017, 5:14:46 PM11/10/17

to Safecast Device Discussions and Support

Hi,

I noticed the changes in the API some time ago including the color scale etc. Does the dose rate calculation still use this method using the CPM parameter in the LOG file:

CPM is converted to µSv/h on a per-device basis.
For bGeigies, the conversion is CPM/350.0

http://safecast.org/tilemap/methodology.html

We had this measurement:

ader_microsvh	time_local	device	device_id	date_time	cpm	pulses5s	pulses_total
0,0828571441	16:05:05	$BNRDD	2847	2017-09-26T14:05:05Z	29	2	266
0,0771428583	16:05:10	$BNRDD	2847	2017-09-26T14:05:10Z	27	1	267
0,0685714296	16:05:14	$BNRDD	2847	2017-09-26T14:05:14Z	24	0	267
0,0685714296	16:05:19	$BNRDD	2847	2017-09-26T14:05:19Z	24	2	269
0,342857148	16:05:24	$BNRDD	2847	2017-09-26T14:05:24Z	120	98	367
0,3457142909	16:05:30	$BNRDD	2847	2017-09-26T14:05:30Z	121	5	372
0,342857148	16:05:35	$BNRDD	2847	2017-09-26T14:05:35Z	120	1	373
0,3485714338	16:05:40	$BNRDD	2847	2017-09-26T14:05:40Z	122	3	376
0,342857148	16:05:45	$BNRDD	2847	2017-09-26T14:05:45Z	120	2	378
0,3457142909	16:05:50	$BNRDD	2847	2017-09-26T14:05:50Z	121	3	381
0,3457142909	16:05:55	$BNRDD	2847	2017-09-26T14:05:55Z	121	2	383
0,3485714338	16:06:00	$BNRDD	2847	2017-09-26T14:06:00Z	122	3	386
0,3600000054	16:06:05	$BNRDD	2847	2017-09-26T14:06:05Z	126	6	392
0,3657142912	16:06:10	$BNRDD	2847	2017-09-26T14:06:10Z	128	3	395
0,3771428628	16:06:15	$BNRDD	2847	2017-09-26T14:06:15Z	132	4	399
0,3771428628	16:06:20	$BNRDD	2847	2017-09-26T14:06:20Z	132	2	401
0,1085714302	16:06:25	$BNRDD	2847	2017-09-26T14:06:25Z	38	4	405
0,1000000015	16:06:30	$BNRDD	2847	2017-09-26T14:06:30Z	35	2	407
0,1000000015	16:06:35	$BNRDD	2847	2017-09-26T14:06:35Z	35	1	408

We use the same formula to calculate the dose rate to be make the QGIS plugin as much consistent with the API as possible.

I also uploaded this sample in a LOG format here:

https://api.safecast.org/en-US/bgeigie_imports/32497

If you compare the API and our table it is clearly visible that the API only has higher dose rate for the only one point (marked red in our table) and the maximal dose rate is much higher than ours (API says 3.53 microSv/h, we have 0.342 microSv/h). So am I right that you now calculate the "API CPM" from 5s pulses data?

eg. CP5s * 12 = CPM, and then again dose rate = CPM/350

(for our maximum it would be 98 * 12 = 1176, 1176/350 = 3.36 so still differs from the API)

or is there more sophisticated approach? It would be useful to know this so we could implement this in next generation of the QGIS plugin too.

thanks

Jan

Kalin KOZHUHAROV (Safecast)

unread,

Nov 10, 2017, 5:46:48 PM11/10/17

to Jan Helebrant, Safecast Device Discussions and Support

On Fri, Nov 10, 2017 at 11:14 PM, Jan Helebrant <jan.he...@gmail.com> wrote:
>
> I noticed the changes in the API some time ago including the color scale etc. Does the dose rate calculation still use this method using the CPM parameter in the LOG file:
>
>> CPM is converted to µSv/h on a per-device basis.
>>
>> For bGeigies, the conversion is CPM/350.0
>

AFAIR, it is not 350, but 334, e.g. see
https://github.com/Safecast/safecastapi/blob/7cbf30add95b9616f0a3d13087e9f9404f78279e/script/ios_query.sql#L14

> I also uploaded this sample in a LOG format here:
> https://api.safecast.org/en-US/bgeigie_imports/32497
>
> If you compare the API and our table it is clearly visible that the API only has higher dose rate for the only one point (marked red in our table) and the maximal dose rate is much higher than ours (API says 3.53 microSv/h, we have 0.342 microSv/h). So am I right that you now calculate the "API CPM" from 5s pulses data?
>
> eg. CP5s * 12 = CPM, and then again dose rate = CPM/350
>
> (for our maximum it would be 98 * 12 = 1176, 1176/350 = 3.36 so still differs from the API)
>

And the exact code I guess is this
https://github.com/Safecast/Tilemap/blob/ca21153293c6647dc515c6798e87877d08d4b43a/bgeigie_viewer.js#L2672

> or is there more sophisticated approach? It would be useful to know this so we could implement this in next generation of the QGIS plugin too.
>

There are always more sophisticated approaches!
But probably not needed. May be think of a more elegant approach?

Cheers,
Kalin.

Richard Wood

unread,

Nov 11, 2017, 12:10:09 AM11/11/17

to Safecast Device Discussions and Support

The below is copied from my bGeiges safecast.txt file, if I understand correctly these are the display setup and other parameters. Seems my Cs137 to Sv number from KitHub was set at 334. I avoid relative dose measurements focusing on counts, we just don't know what isotope the meter is measuring.

nm=Safecast
did=3023
gt=0
gm=0
cpmf=334
cpmn=Cs137
bqmf=37
bqmn=Cs137
al=150
tz=-8
cn=US
st=0
ss=0
sh=100
sm=0
cw=60

Jan Helebrant

unread,

Nov 11, 2017, 8:41:08 AM11/11/17

to Kalin KOZHUHAROV (Safecast), Safecast Device Discussions and Support

dear Kalin,
thanks for your reply.

> AFAIR, it is not 350, but 334

In fact we use 334 as I found in our documentation later. However if you
look in this table:

https://docs.google.com/spreadsheets/d/1BDWza6TrPZszwqCfnZtBj9tOEhFmyGAWhoU0bYM_BJ8/edit?usp=sharing

I made for this example you see that if you use CPM from the LOG file to
calculate the dose rate you have some "fade out" effect (the higher dose
rates continue some points althought the CP5s values are low) although
there is only one elevated point. The API now does not have this effect
so it looks like CP5s is used to calculate the CPM and then the dose
rate and LOG CPM is used only if the CP5s is zero. The conversion
coefficient seems to be still the same - 0.0029940119760479.

Jan

Dne 10.11.2017 v 23:46 Kalin KOZHUHAROV (Safecast) napsal(a):

Jan Helebrant

unread,

Nov 13, 2017, 6:49:19 PM11/13/17

to Safecast Device Discussions and Support

OK,

I did some tests with randomly selected data points in two LOG files and it looks like the dose rate for the API is now using CPM calculated from the CP5s values.

Fukushima data examples		https://api.safecast.org/en-US/bgeigie_imports/18170			uploaded	2015-03-25
All taken from API:
LOG data		API calculated		calculation check		calculation from LOG CPM
CP5s	CPM	CPM	dose rate microSv/h	CPM from CP5s	dose rate microSv/h	dose rate microSv/h
56	604	673	2,01	672,00	2,0120	1,8084
93	1572	1118	3,35	1116,00	3,3413	4,7066
10	84	120	0,36	120,00	0,3593	0,2515
11	227	132	0,40	132,00	0,3952	0,6796
7	65	84	0,25	84,00	0,2515	0,1946
61	780	733	2,19	732,00	2,1916	2,3353
713	5527	8696	26,04	8556,00	25,6168	16,5479
787	5992	9615	28,79	9444,00	28,2754	17,9401
594	7098	7225	21,63	7128,00	21,3413	21,2515
17	1136	204	0,61	204,00	0,6108	3,4012
775	3455	9466	28,34	9300,00	27,8443	10,3443
1274	4538	15741	47,13	15288,00	45,7725	13,5868

CPM from CP5s = CP5s * 12
dose rate = CPM * 0,0029940119760479

Damětice data examples		https://api.safecast.org/en-US/bgeigie_imports/32105			uploaded	2017-10-22
All taken from API:
LOG data		API calculated		calculation check		calculation from LOG CPM
CP5s	CPM	CPM	dose rate microSv/h	CPM from CP5s	dose rate microSv/h	dose rate microSv/h
3	294	36	0,11	36,00	0,1078	0,8802
11	263	132	0,40	132,00	0,3952	0,7874
32	444	384	1,15	384,00	1,1497	1,3293
60	604	721	2,16	720,00	2,1557	1,8084
82	757	986	2,95	984,00	2,9461	2,2665
67	794	805	2,41	804,00	2,4072	2,3772
35	516	420	1,26	420,00	1,2575	1,5449
40	515	480	1,44	480,00	1,4371	1,5419
12	97	144	0,43	144,00	0,4311	0,2904
8	95	96	0,29	96,00	0,2874	0,2844
7	95	84	0,25	84,00	0,2515	0,2844
9	93	108	0,32	108,00	0,3234	0,2784

CPM from CP5s = CP5s * 12
dose rate = CPM * 0,0029940119760479

whole spreadsheet:

https://docs.google.com/spreadsheets/d/1BDWza6TrPZszwqCfnZtBj9tOEhFmyGAWhoU0bYM_BJ8/edit?usp=sharing

in my other LOG file:

https://api.safecast.org/en-US/bgeigie_imports/32497

there are also points with CP5s = 0 and in that case "LOG CPM" is used for dose rate calculation. I has to be also somewhere in the API code but I am not programmer. I hope next generation of QGIS plugin will use this approach too.

Jan

Reply all

Reply to author

Forward