Edzer, You are right about river networks.
I do think that the uni-directionality of flow is not necessarily a problem in terms of what can inform what. Statistically, a downstream station can inform an upstream station because what is downstream is there because of what happened upstream, and the same for past and future. While there may not be a *physical* link from future to past, there will still be a *statistical* link from future to past as long as there was a statistical link from past to future. [That is the idea in inverse modeling for example (which are a combination of physical and statistical modeling)]. Intuitively, kriging backwards might have to be done according to a different model than forward, but, I am open to surprises. It is worth a try, anyway. Noémi -----Original Message----- From: Edzer J. Pebesma [mailto:[EMAIL PROTECTED] Sent: Thursday, February 24, 2005 5:19 PM To: Noemi Barabas Cc: oscar garcia; ai-geostats@unil.ch Subject: Re: [ai-geostats] Kriging along streams Noemi, the idea of transforming to a one-dimensional situation, distance along river, was the first one I got. Then, I rejected it because it will not work for river networks; it will for a single sections without joints though. Space-time modelling may be complicated by the fact that there will be a strong correlation with future observations downstream, but not with observations upstream because water flows down; this may call for another translation of space-time using flow speed to get to a space-time coordinate system where the usual models make sense. -- Edzer Noemi Barabas wrote: > Dear Oscar, > > It sounds like you need kriging in a single dimension. If so, I think what > you need to do is a kind of coordinate transformation, for example using > river mile as your distance vector. You would also need to pay attention to > the dates of the samples. Variation in time might obscure spatial trends > along the river's course if you just have a few sampling events at different > times for the stations. > > If you have data in space as well as time, there are several methods for > space-time interpolation. Constructing correlograms for data vs time on > paired stations could give you an idea of what constitutes a data pair for > kriging purposes in terms of the time lag between them. Methods will differ > depending on whether it is the time or space dimension that is more densely > sampled. > > References on techniques include the following: > > For straightening rivers (which might be more involved than you need): > > Barabas, N., Goovaerts, P. and P. Adriaens. 2001. Geostatistical assessment > and validation of uncertainty for three-dimensional dioxin data from > sediments in an estuarine river. Environmental Science & Technology, 35(16): > 3294-3301. > > For space-time data: > > Pebesma and de Kwaadsteniet (1997) Mapping spatial and temporal variation of > groundwater quality in the Netherlands. In GeoENVI - Geostatistics for > Environmental Applications, eds., Soares et al, pp. 139-151. Kluwer > Academic Publishers, Netherlands. > > Kyriakidis and Journel (1999) Geostatistical space-time models: A review. > Mathematical Geology (not sure which volume etc.) > > Heuvelink et al (1996) Spatio-temporal kriging of soil water content. In > Geostatistics Wollongong '96, eds Baafi and Schofield. pp. 1020-1030. > Kluwer Academic Publishers,Netherlands. > > > Noémi > > > > -----Original Message----- > From: oscar garcia [mailto:[EMAIL PROTECTED] > Sent: Thursday, February 24, 2005 11:56 AM > To: ai-geostats@unil.ch > Subject: [ai-geostats] Kriging along streams > > > Dear list members, > > I want to perform kriging of water quality indicators > measured in stations along a river course, so Can you > help me with references about kriging along streams in > a river basin?? is there software to perform this kind > of analysis??? > > Thanks in advance for your help, > > Oscar Javier García-Cabrejo > Master in Hydrosystems, Civil Engineering Department > Pontificia Universidad Javeriana > Bogota, Colombia > South America > > _________________________________________________________ > Do You Yahoo!? > La mejor conexión a internet y 25MB extra a tu correo por $100 al mes. > http://net.yahoo.com.mx > > > > > ------------------------------------------------------------------------ > > * By using the ai-geostats mailing list you agree to follow its rules > ( see http://www.ai-geostats.org/help_ai-geostats.htm ) > > * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] > > Signoff ai-geostats -- Edzer
* By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats