|
|||||||
|
29 October 2013 NSA Boundless Informant Images Explicated Related: 2013-1457.htm NSA Spied 125 Billion Phone Calls in a Month October 24, 2013
Date: Mon, 28 Oct 2013 23:36:25 -0700 (PDT) Due to international press coverage, particularly in Italy and the Middle East, I have decided to enclose the images from BOUNDLESSINFORMANT. And explain in more detail what I know about them. Heatmaps: The three "heatmaps" show the data represented 3 different ways: DNR (telephony metadata), DNI (internet metadata), and Aggregate (the total of the two). DNI and DNR are actually broader terms meaning Digital Network Intelligence, and Dialed Number Recognition). The BOUNDLESSINFORMANT tool however, is only capable of displaying metadata. When: In all heatmap images there is a declassification date (20320108), from which you can calculate the date of the capture (January 8, 2013). This is only legible in the image published in Le Monde. How Long: Also on all heatmaps, a gray box with green text reading "Top 5 countries (last 30 days from today) is shown. Again only legible in the Le Monde image. How Many Records: On all heatmap images, the United States is always shown. In the "Top 5 countries" frame there are also DNI and DNR Totals by country, a grand total is given in the "Overview" frame. Because the "top five" list is represented three different ways, we wind up with more than five countries once we have examined the images. The heatmaps have appeared in various papers: Le Monde, The Guardian, Der Spiegel, and The Hindu. The Le Monde images are of excellent quality, the others are not as sharp. The names of the countries can be determined, in part, by looking at the map. Not all of the digits of all the numbers can be determined with certainty, so I have only attempted to read the first few. Some countries and their totals appear on multiple image, which is an excellent help when dealing with low-quality images. Bar Graphs: The bar graphs show daily totals, the last 30 days, from December 10, 2012, to January 8, 2013 (which just happens to match the map view). A graph of France has been produced by Le Monde, and 30 Aggregate (DNI+DNR) totals day totals are given 2 different ways: The "Most Volume" box, and the sum of the "top 5 techs" boxes. In the case of France, there is no DNI to deal with, so the Aggregate reflects only DNR. One might think i could save a lot of time by not measuring this graph, but in fact I need it as a check to determine if the bars are in fact precisely scaled, and to validate the accuracy of my methods. Because this is a high quality image, it allows for a higher degree of precision than the other images. The remainder of the graphs were published in Der Spiegel, including a second copy of the French map (with the bottom bit cropped). The German graph, which Der Spiegel annotated, shows the bottom bit. But it is not as helpful as the French one. It does not allow me to check my DNI and DNR totals individually. And the sum of the top 5 techs is less than the sum of the 2 "most volume" SIGADs. I can still check my Aggregate sums against the sum of the 2 SIGADs. My aggregate total for Germany works out to be 1.4 million lower than the known true aggregate total. In addition, there are graphs for Italy, Spain, and the Netherlands. A small partial graph, belonging to Germany was also published in the Guardian, for the XKEYSCORE article. Der Spiegel mentions, but does not show an additional graph for Poland. "In Poland, which is also under surveillance, the numbers varied between 2 million and 4 million in the first three weeks of December". Without seeing the Poland image, we can still estimate a range (2 million*21 days=42 million; 4 million*21 days=84 million) It doesn't state whether this is DNI, DNR, or Both, nor is it precisely clear why there are 21 days described, and not 30. Greenwald has written in El Mundo that the total for Spain is 60.5 million. I am unable to determine whether he has gotten his number from my analysis (which appeared well before his article) or if he has a separate total. If it's the latter, it speaks well for my tallies in Italy and Netherlands. Methodology: Simple enough: measure the scale on the graph, measure the bars. I used GIMPs Measuring Tool to count pixels. Enter the pixels into a spreadsheet, and let it calculate the approximate number. Accuracy: this depends on the scale, one pixel for measurement, multiplied by the scale, which varies from graph to graph. Don't be confused because the scaled number appears to be precise, it is a calculation. In general terms, you will get no more than 2 significant figures. Clarifications to various media reports: * 1 phone call may generate multiple metadata records. For reasons I have already mentioned. Nor is it necessarily the case that each call has had the contents recorded. That will depend on other factors: for example whether the person has been "tasked" (ie targeted), and which infrastructure the call has been routed through. All calls routed through the UK will pass through Tempora; this means the the CONTENT will indeed be stored for about 3 days, and the metadata for 30. All calls routed through the United States will have the metadata recorded for about five years, and for those unfortunate enough to encounter AT&T's Hemisphere project, possibly as far back as the mid 1980s. I do not know how much content will be stored, and for how long in that case. Duncan Campbell writes in the Independent that a facility similar similar to and associated with Tempora exists somewhere in the Middle East, but does not specify where. Nor is it clear precisely how he obtained Snowden information, other than it wasn't through Mr Snowden himself. I SPECULATE he may have gotten it from Suddeutsche Zeitung, which had published a similar story. What happens in the rest of the world, I do not have information on and decline to speculate. Any people who have been "tasked" could have their calls potentially memorialized indefinitely. * I am not Cryptome * I am not a second NSA mole, I am an independent analyst, without portfolio. I work for no one and have no secret knowledge beyond the media reports. I have compiled this analysis on my own time. * There is not "no evidence" I am particularly concerned by this statement, by Italian intelligence officials. These graphs should have prompted pan-European concern from the moment they appeared on Der Spiegel. Somebody in each intelligence agency should have done precisely what I have, and sooner. If they have not done so, it does not bode well for their capabilities and credibility as intelligence agencies. They should have been checking this through Diplomatic channels, and via their own means. For the sake of their credibility, I hope they are simply dissembling. * I would appreciate if any political editorializing is attributed to the appropriate source, and not either Cryptome, or me. My goal in this forum is simply to inform, and give correct information, the only true basis for correct decisions. * For my one bit of political editorializing, I would very much appreciate it if this analysis does not lead me to be put on the No Fly List, or become "tasked" myself. And need to remind appropriate parties just where this information comes from in the first place. Additional Sources: http://www.theguardian.com/world/2013/jun/06/nsa-phone-records-verizon-court-order http://www.theguardian.com/uk/2013/jun/21/gchq-cables-secret-world-communications-nsa http://www.nytimes.com/interactive/2013/09/02/us/hemisphere-project.html?_r=0 http://www.elmundo.es/espana/2013/10/28/526dcbad61fd3d07678b456b.html
|
![[Image]](nsa-boundless-informant-images-01.jpg)
![[Image]](nsa-boundless-informant-images-02.jpg)
![[Image]](nsa-boundless-informant-images-03.jpg)
![[Image]](nsa-boundless-informant-images-04.jpg)
![[Image]](nsa-boundless-informant-images-05.jpg)