deletions | additions       diff --git a/Results_and_discussion_Matched_Filter__.md b/Results_and_discussion_Matched_Filter__.md  index 0d21381..f82d542 100644  --- a/Results_and_discussion_Matched_Filter__.md  +++ b/Results_and_discussion_Matched_Filter__.md 
...
# Results and discussion  ##Matched Filter detections  Using all 637 template events, we were able to detect more than 20,000 additional events from 1 January to 20 November, 2015. A large majority of these (>12,000) were located in Rotokawa while slightly less than 4000 events were located in Ngatamariki. Ngatamariki (Figure \ref{Figure2}).  These numbers conform to the 3:1 Rotokawa-Ngatamariki seismicity ratio mentioned earlier (e.g. \cite{Sherburn&Bourguignon_2015}, \cite{Sherburn&Johnson_2015}). Detections have been clustered into three basic groups: Rotokawa, northern Ngatamariki and southern Ngatamariki, based upon historic seismicity. The cumulative detections for each of these groups is plotted with time for the study period along withwell  pumping rates for pertinent injection wells (Figs. \ref{Figure3} \ref{Figure4} \ref{Figure5}). At Rotokawa Rotokawa,  initial inspection seems to show no obvious correlation between rates of injection and rates of matched filter detection (Figure \ref{Figure3}). Instead, the cumulative rate of detection for all Rotokawa template events is roughly constant throughout 2015 with small, abrupt increases in detection rate which do not appear correlated with large changes in rate of injection. Notably, the annual plant maintenance shutdown in July 2015, during which all injection was halted, did not produce an increase in the rate of seismicity as has been observed in the past (Steve Sewell, personal communication). In Northern Ngatamariki, the same observations hold true (Figure \ref{Figure4}). The cumulative detection rate remains roughly constant throughout the year, with no obvious increase during plant shutdown and or  restart. In southern Ngatamariki, however, we do see an obvious increase in cumulative detection rate at the time of plant shutdown (Figure \ref{Figure5}). Notably, however, short, 1-2 day increases in cumulative detection rate similar to the increase associated with plant shutdown are also present during parts of the year during which there are no notable, large-scale changes in injection rate across the field. Specifically at southern Ngatamariki, we see a series of very abrupt detection rate increases at five separate times during the year (Figure \ref{Figure5}). This may represent swarm-like behaviour occurring along the Aratiatia Fault Zone (Figure \ref{Figure1})  which has generated microseismicity in the past and which provides a direct link for fluid travelling traveling  between injection well NM10 and production well NM5 \cite{buscarlet2015reservoir}.