Latest Contributions by EricKrause

‎01-26-2016

Thank you for this feedback. As others have noted, Interpolate Points does not use IDW for the interpolation. Instead, it uses Empirical Bayesian Kriging (EBK), which we consider to be our best automatic interpolation method. We will add extra information about the particular EBK settings that are used in Interpolate Points to the REST API documentation in the near future, but for now, this is what you need to know. Most parameters from EBK are controlled by the "Optimize for" parameter in Interpolate Points. This parameter has three setting: Speed, Accuracy, and an unnamed default option that I will call "Medium". If you choose the "Speed" setting, EBK will use the following parameters: Data transformation type: NONE Semivariogram model type: POWER Maximum number of points in each local model: 50 Local model area overlap factor: 1 Number of simulated semivariograms: 30 Number of neighbors: 8 If you choose the "Medium" (default) setting, EBK will use the following parameters: Data transformation type: NONE Semivariogram model type: POWER Maximum number of points in each local model: 75 Local model area overlap factor: 1.5 Number of simulated semivariograms: 100 Number of neighbors: 10 If you choose the "Accuracy" setting, EBK will use the following parameters: Data transformation type: EMPIRICAL Semivariogram model type: K_BESSEL Maximum number of points in each local model: 200 Local model area overlap factor: 3 Number of simulated semivariograms: 200 Number of neighbors: 15 You can learn more about these parameters here. Let me know if you have any other questions.

‎12-17-2015

The sill and range do not refer to any particular elevation value (in fact, they apply to all elevation values). These numbers tell you the distance at which two points are no longer correlated (the range), and the expected squared difference in their elevation values (sill). The semivariogram plots the squared difference in elevation between pairs of points that are a given distance apart. It only operates on the difference in elevation values, not on the actual elevation value. For example, imagine two points that are 2000 meters apart. One has an elevation of 1000 meters, and the other has an elevation of 1100 meters. The squared difference in elevation value is 100^2=10000, so this pair would get plotted at (2000, 10000) in the semivariogram. Now, imagine another pair of points that are also 2000 meters apart, but their elevation values are 100 meters and 200 meters. Their squared difference in elevation is also 10000, so this pair would also be plotted at (2000, 10000) in the semivariogram. Hopefully you can see that you cannot extract an elevation value where the the semivariogram reaches the sill because semivariograms are not constructed to answer that kind of question.

‎12-09-2015

Windows 10 is indeed supported for ArcGIS 10.3.1, but Error 999998 does not imply that the error is due to the operating system. If possible, can you send your data and workflow to ekrause@esri.com? I should be able to figure out what is going wrong. Otherwise, you can contact Esri support services to have your data handled securely.

‎11-10-2015

Without seeing your data, it's hard to give specific recommendations. However, cokriging and regression kriging work in completely different ways, so you shouldn't expect their results to align perfectly. I don't know if the difference is big enough to be worried about, but you shouldn't expect the maximums to be the same. About the log transformation, again, it's hard to recommend whether or not to transform without seeing the data. But if you think a log transformation is needed, this is how you would proceed: Take the log of the dependent variable at the very beginning. Run OLS on the transformed variable, interpolate the explanatory variables and the residuals, and plug the results into the equation as before. This will give you predictions for the log of the data value. You would then take the anti-log (e^predictions if you used natural log) to get predictions on the original scale.

‎11-05-2015

If your organization has permissions to download prerelease versions of ArcGIS, you will be able to download them through my.esri.com.

‎11-05-2015

This is what I mean by plugging the the rasters into the OLS equation: When you run OLS, you will get coefficients for the explanatory variables along with an intercept value. For example, let's say that the precipitation variable gets a coefficient of 2, elevation gets a coefficient of 3, and the intercept term is -4. You would interpolate elevation, precipitation, and residuals to rasters. You would then apply the following formula to get your prediction surface: Prediction = -4 + 2 * (precipitation) + 3 * (elevation) + (residual) The Raster Calculator tool is the easiest way to apply this formula to all raster cells. For example, in a location where the precipitation is 500, elevation is 1000, and the residual is 50, the prediction would be: Prediction = -4 + 2 * 500 + 3 * 1000 + 50 = 4046 So, your prediction at that location would be 4046. I will look into how you can download ArcGIS Pro 1.2 beta. It will be available as part of the ArcGIS for Desktop 10.4 release, so if you are on a maintenance contract that has this product available, you will be able to download the beta.

‎11-04-2015

There is an issue in your third step. I'm assuming that your dependent variable and independent (explanatory) variables are coming from fields in a point feature class. In this setup, there is no "correct" way to produce a prediction surface. The explanatory variables must be measured in every location where you want to make a prediction. I understand that this is often impractical, and what is often done in this case is to interpolate the explanatory variables. It is not a good idea to interpolate the predicted values because you know that they depend on the explanatory variables. However, when you interpolate the explanatory variables, all subsequent analysis will assume that the interpolated values are measured values when, in fact, they are predictions and have associated error. In your case, you need to calculate the OLS equation from the input points, then interpolate the independent variables and the residuals, and plug the results into the OLS equation to get the predicted values. Regarding cross validation, only the residuals (error terms) are needed. When you are interpolating the residuals, note the cross validation statistics. These are what you should report as the cross validation statistics for the regression kriging model. However, as I alluded to in my first post, we are releasing "EBK Regression Prediction" in ArcGIS Pro 1.2, which does a regression kriging interpolation using Empirical Bayesian Kriging. It will very shortly be available as a Beta through Esri's customer care website. I am not sure exactly where to go to download it, but I can find out if you are interested.

‎08-31-2015

No, you cannot control the maximum (or minimum) value of the interpolation. Whatever the maximum is (if there even is a maximum) will come from the kriging equations. These equations are constructed based on all the provided parameters, and the maximum/minimum will be whatever the equations say they are. In fact, if we could detect the maximum and minimum, my life would be a lot easier. Currently, we symbolize geostatistical layers based on the min/max of the input data, not the actual interpolation. If we knew the min/max of the interpolation, we could symbolize the geostatistical layer based on these values. If you have ever noticed that geostatistical layers often look very different after being exported to raster, it is because of this issue. If we could detect the min/max of the geostatistical layer, we could make the symbology closely match the raster.

‎08-17-2015

The primary dataset almost always has much more impact than the cokriging dataset, so if you're going to prioritize any of them, Var1 - Var1 should be the priority.

‎08-14-2015

Correction to something I said earlier. We apparently do support negative cross-covariances (I actually did not know this). The covariances of Var1 - Var1 and Var2 - Var2 do have to be positive, but the cross-covariance between them (Var1 - Var2) can indeed be negative. The idea is that each dataset has to be positively spatially correlated with itself, but the cross-correlation between them can be negative. This will be useful in cases where both the primary and cokriging dataset are spatially correlated (again, to themselves), but the two variables have an inverse relationship, ie, when one is large, the other tends to be small. As for other diagnostics, the RMS is just one crossvalidation statistic. The others are described in this help document: Cross Validation—Help | ArcGIS for Desktop You should also look at the crossvalidation summary statistics, the locations of largest/smallest errors, and the graphs on the final page of the Wizard.

‎08-14-2015

The optimize button does not do a complete optimization. For example, the choice of semivariogram model is not optimized. The default model is "Stable," and the optimize model button will find the range, nugget, and sill that minimize the RMS for the Stable model. If you change, for example, to K-Bessel, pressing the optimize model button will find different optimal parameters for the K-Bessel model, which may or may not be better than the optimal Stable parameters. But again, the model with the lowest RMS is not necessarily the best model; there are lots of other diagnostics that you should pay attention to. If those other diagnostics do not look good, the model may need some manual changes. If you want to manually specify the parameters for the three models, you change them directly with the parameters on the right side of the wizard. Changing Var1 - Var1 to Var2 - Var2, for example, just changes which graph you are looking at, but you can control the parameters of all three models no matter which one you are currently looking at. You don't need to restart the Wizard after each change. Hitting the optimize model button will set all parameters to their optimal values, no matter what their current values happen to be. If you want to get back to the original default values (before you pressed optimize), you either need to restart or click Back a couple times, then Next a couple times, and they will revert to their non-optimal defaults. With regards to Figure 11 in your link, you're looking at a covariance view of the semivariogram. You can get your graph to look like this by changing the "Variable" setting on the top-right of the wizard to "Covariance". This just just a different view of the same thing. Instead of making a graph of squared differences (the semivariogram), it makes a graph of covariances, and the idea is that points that are close together are more correlated (ie, have a larger covariance) than points that are further apart. After a particular distance (the range), the covariance becomes 0, which means that points that are further apart than the range are considered spatially independent. You'll also notice that the blue covariance curve never goes below zero, which is what a true "negative covariance model" would look like. The general idea is that the semivariogram plots how different points are, and the covariance view plots how similar they are. So, points that are farther away will have larger semivariances, but they will have lower covariances.

‎08-13-2015

In Cokriging, if you press the optimize model button, the software will simultaneously optimize the semivariogram for the primary dataset (Var1 - Var1), the semivariogram for the cokriging dataset (Var2-Var2), and the crosscovariance between them (Var1 - Var2). If you do not want to rely on the optimize model button, you can manually specify the parameters for all three covariance models. What the optimize model button is actually doing is finding the set of parameters (for all three covariance models) that minimize the root-mean-square (RMS) crossvalidation error. However, the RMS is just one of many diagnostics, and it isn't uncommon for the model with the lowest RMS to fail at other diagnostics. Negative cross-covariances cannot be modeled in ArcGIS. All covariance models supported in ArcGIS assume positive spatial correlation that diminishes over distance.

‎06-10-2015

If you are using ArcGIS 10 or 10.1, you will need to pass the datasets using strings. The GeostatisticalDatasets object will not work because it does not exist in 10 or 10.1. For an example of the syntax of the string, see the code samples at the bottom of the ArcGIS 10 help topic: http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#/Create_Geostatistical_Layer/003000000015000000/

‎06-05-2015

This workflow is really only suggested when you know that the new datasets will have the same general structure. Applying the same parameters to datasets that are different (ie, the presence and absence of trends) is not a good idea. As for trend removal, the idea is to first do a very basic polynomial interpolation to remove the general trends, then you do kriging on the autocorrelation that is left over. The idea is that there is a general trend in the data, but there is also correlated variability around that trend. Trend removal takes care of the first part, and kriging takes care of the second part. In essence, you are interpolating twice, but the first time is only for general trends in the data.

‎06-05-2015

This is difficult question to answer in an online setting. The differences are very technical and mathematical, but the key difference is that the covariance view requires knowledge of the mean value of the kriging surface, and the semivariogram does not. If you are unsure which one to use, I would suggest the semivariogram. Both methods are designed to estimate covariance matrices that are used in the kriging equations, but they both do this estimation in different ways, and their "optimal" parameters will be different (though hopefully not very different).

Online Status	Offline
Date Last Visited	‎02-25-2026 06:39 PM

My Ideas

Latest Contributions by EricKrause

Re: Interpolating arcgisonline vs ArcGIS Spatial Analysis

Re: semivariogram analysis using geostatistical tool in arcgis.

Re: Windows 10 Problems

Re: Regression kriging

Re: Regression kriging

Re: Regression kriging

Re: Regression kriging

Re: Geostatistical analyst layer

Re: questions about cokriging model selection

Re: questions about cokriging model selection

Re: questions about cokriging model selection

Re: questions about cokriging model selection

Re: loop through geostatistical layers in a folder

Re: Geostatistical analyst layer

Re: Geostatistical analyst layer

Re: Why exporting EBK 3D to voxel layer change org...

Re: Access geostatistical layer created using ArcP...

Re: Why exporting EBK 3D to voxel layer change org...

Re: K-Bessel Semivariogram Equation

Re: Kriging Model Types

R-ArcGIS