ChangeLog

Changes from version 1.2 to 1.2.1   3 Oct 2017
	* added vignette to provide usage examples

Changes from version 1.1.1 to 1.2   9 Jan 2013
	* added VPRM parameter estimation functionality.  Currently this
	means that DEoptim package (differential evolution optimization)
	is now required.  I may make this a "Suggests" package in the
	future.
	* (estimateVPRMPars.R) added optimization "windows" for VPRM
	parameters lambda, PAR_0, alpha, and beta as optional parameters
	to the functions estimate_VPRM_pars and optimizeVPRM_DE.

Changes from version 1.1 to 1.1.1   20 Dec 2013
	* added VPRM_parameters dataset and documentation

Changes from version 1.0 to 1.1   18 Dec 2013

	* added Park_Falls dataset
	* added example code for all functions
	* (vprm.R) removed getEVI.  While I don't see anything wrong with
	the arithmetic in the calculation, the EVI values it produces
	don't look very good.  I wrote it before realizing there is a
	MODIS EVI product (e.g. MOD13Q1).  Better to just use the MODIS
	product.
	* (vprm.R) changed the "tower" argument for NEE and GEE to
	"driver_data".  I think this more accurately reflects its
	contents, which do not come exclusively from eddy covariance
	towers (some driver data come from satellites or, in some cases,
	reanalysis products).
	* changed the package name from "VPRMModel" to "VPRMLandSfcModel".  The name
	VPRMModel reflects an earlier state of the R code where the
	functionality was divided between two packages, VPRMModel and
	VPRMFramework.  I have eliminated VPRMFramework toward the goal of
	having everything in one package.  Now VPRMModel is redundant
	(because the "M" in VPRM already abbreviates "model").  I think
	that VPRMLandSfcModel, while perhaps similarly redundant with
	"Model", strikes a good balance between brevity, describing the
	model's purpose at a glance.  I considered naming the package
	"VPRM", but I thought that might lead to confusion between the
	model itself and the R package implementation.