How to interpolate ibutton transect temperature data ArcGIS

[JoeBanker]

I have three transects on opposing slopes. Each transect has ibutton data loggers set to record temperature at hourly intervals. The loggers are spaced at 20m altitudinal intervals from 1550 - 1810m (west-facing slope) and 1550 - 1690m (east-facing slope). Each transect has three ibuttons per altitudinal interval for replication and averaging purposes.

I have the average temperatures for each altitudinal interval (average of three ibuttons for each elevational interval) and have split the data according to seasons.

I would like to interpolate the averaged seasonal temperatures of to create a temperature layer (interpolate to a DEM) which I would then want to use as a covariate for occupancy modelling of a vulnerable species of lizard.

The other problem is that in the south of my study site the elevation is ca. 1450m (see attached map for context) 

for which I have no temperature data so I would have to interpolate outside of my known data points (if that makes sense).

How can I do this in ArcGIS (if it is possible to do)?

[DanPatterson_Retired]

Steve Lynch‌ … for you

interpolating along the transects would be fine...

interpolating between the transects, maybe (Steve had suggestions on another thread that I can't locate right now...

interpolating into the void beyond the data values, I will leave to Steve

[SteveLynch]

Dan Patterson‌, not an easy one to answer 🙂 and the link you could not find is to Interpolation of Data Collected along Lines

I spoken with some of my colleagues and the consensus is;

• be careful of predicting outside the data's extent (i.e. extrapolating)
• possible solution is to use EBK Regression Prediction which is available in ArcGIS Pro;
  • explore the option of supplying measurement values, if not known you could use data replication to estimate them
  • use elevation and additional covariates like roughness index, aspect, distance to something  

Steve

[JoeBanker]

Thank you very much for your input. I'm a bit of a GIS newbie, could you perhaps explain what you mean by "supplying measurement values and data replication"?

Also to create the interpolated temperature layer, should I pool the temperature data together from the opposing slopes?

I apologise if my questions are a bit redundant I'm just a bit confused about the workflow that I should follow.

[DanPatterson_Retired]

Thanks Steve... I am a little more reserved about the first consensus item

• don't ever predict outside the data's extent (i.e. extrapolating)

unless you have a follow-up study to indicate that the extrapolation was prudent... seen too many non-spatial and spatial extrapolations go bad I suppose

[JoeBanker]

Thanks for your input Dan.