Thank you very much, sir. I have been watching some of your videos; they are helpful for my thesis. I look forward to watching more of yours.
@yosraelmaghur1109
6 ай бұрын
Thank you very much for this video, it was very clear and well organized. I really appreciate sharing your knowledge with us.
@normjones
6 ай бұрын
Thank you for the kind feedback.
@prattmk2010
6 ай бұрын
Great presentation. Is (observation weight = 1/variance) or is (observation weight = 1/standard deviation)? I thought it was the latter, but on the "Error Norms, ctd." slide, it seems to be the former. Thanks
@normjones
6 ай бұрын
Good question. There are multiple ways to define the observation weights. See the MODLOW PES documentation here: water.usgs.gov/nrp/gwsoftware/modflow2000/MFDOC/pes.html And scroll down to Data Set 10 and look at the STATP and STAT-FLAG variables. GMS uses the STAT-FLAG = 1 option where the weight = 1/variance and the variance = std dev^2. Therefore the weight = 1/(std dev)^2. This makes it possible to define and think about weights in terms of the interval and confidence, which is more intuitive than standard deviation.
Пікірлер: 5