This demonstration contains three modules that can be accessed from the tabs at the top.


Time series: Data points presented over time together with future estimates


Density plots: Daily vounts are displayd in a dual plot.


Spatial data: Data points plotted on a map in time series format.




Throughout the app the following definition is very important: Brightness is the temperature of the fire pixel measured in Kelvin


Introduction

The following app is used to demonstrate RiskCedes time series data analysis tools.

This data set was provided by the University of Maryland and NASA FIRMS operated by NASA/GSFC/ESDIS with funding provided by NASA/HQ NASA Near Real-Time and MCD14DL and MCD14ML MODIS Active Fire Detections (TXT format). Data set. Available on-line here.

The data used is all fires in South Africa between 1 Jan 2012 and 31 Mar 2015.








Time series

An interactive time series plot with predictive values.


Click and drag to zoom in (double click to zoom back out).




The left hand side is a line plot of the actual number of fires per day..

The right hand side is a histogram of the number of days with a specified number of fires (on the y axis).



Histogram of daily number of fires

A dual plot of daily fires.




The following set of plots investigate spatial properties of fires over time.

The plots are split up per season, see below.
Not all seasons have the same duration, so some may include more fires than others due to longer time exposure, but as rainfull is linked to season, this assumed to be the most significant factor.

Summer: From October to February, 5 month duration

Autumn: From March to April, 2 month duration

Winter: From May to July, 3 month duration

Spring: From August to September, 2 month duration


Download the map

Hot zones

A map of the hot zones over time.