| Title: | Analysis and Prediction of Tides |
|---|---|
| Description: | Tidal analysis of evenly spaced observed time series (time step 1 to 60 min) with or without shorter gaps using the harmonic representation of inequalities. The analysis should preferably cover an observation period of at least 19 years. For shorter periods low frequency constituents are not taken into account, in accordance with the Rayleigh-Criterion. The main objective of this package is to synthesize or predict a tidal time series. |
| Authors: | Moritz Mueller-Navarra [aut, cre], Sylvin Mueller-Navarra [aut] ((2019) <https://doi.org/10.5194/os-15-1363-2019>), Andreas Boesch [ctb] ((2019) <https://doi.org/10.5194/os-15-1363-2019>) |
| Maintainer: | Moritz Mueller-Navarra <[email protected]> |
| License: | GPL-3 |
| Version: | 0.0.5 |
| Built: | 2024-10-15 04:45:20 UTC |
| Source: | https://github.com/moritzmn/tidecurves |
Builds the xdesign.matrix by calling Funcs. For internal use.
BuildDesign(tdiffa, numma, numme)BuildDesign(tdiffa, numma, numme)
tdiffa |
The difference in days as double which stems from the analysis period. |
numma |
The transit number (start). |
numme |
The transit number (end). |
Returns a matrix
Builds a TideCurve model of class "tidecurve".
BuildTC( dataInput = NULL, otz = 1, astime, asdate, aedate, aetime, km = -1, mindt = 30, keep_data = FALSE )BuildTC( dataInput = NULL, otz = 1, astime, asdate, aedate, aetime, km = -1, mindt = 30, keep_data = FALSE )
dataInput |
A data frame with the columns observation_date, observation_time and height. See attached data for correct formats. |
otz |
The time zone of the observations |
astime |
A string indicating the time you want the analysis to start with. Format: "hh:mm:ss" |
asdate |
A string indication the date you want the analysis to start with. Format: "yyyy/mm/dd". |
aedate |
A string indication the date you want the analysis to end with. Format: "yyyy/mm/dd". |
aetime |
A string indicating the time you want the analysis to end with. Format: "hh:mm:ss". |
km |
The number of nodes between two consecutive mean moon transits. Shall be less or equal to: round(1440 [min] / time step [min]) Example: Time step 5 min: Use km = 288 or even smaller. Leave on default (km = -1) and supply mindt, when unsure. |
mindt |
Observation time step in [min]. Default is 30. |
keep_data |
Indicating whether you want to keep the data for computing residuals later. Default is FALSE which keeps the model footprint small. |
A model of class tidecurve, which is a list.
## Not run: BuildTC(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2015/12/31", aetime = "23:30:00") ## End(Not run)## Not run: BuildTC(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2015/12/31", aetime = "23:30:00") ## End(Not run)
Returns predictor vector for design matrix from 39 astronomical angular velocities.
Funcs(tdiff, xi)Funcs(tdiff, xi)
tdiff |
Length of input time series. |
xi |
Transit index |
A list with the selected angular velocities, their ranks and the predictor vector (Values between -1, 1).
This function computes the residuum of the computed lunar and solar synthesis and the observed data
ResCurve(tcData, obsData)ResCurve(tcData, obsData)
tcData |
The results from TideCurve or BuildTC + SynTC. Warning: The synthesis period must overlap with the analysis period. Must be a data.table object. Please see examples. |
obsData |
The observation data with the columns observation_date, observation_time and height. See attached data for correct formats. |
A list with two data.tables with the joined data input and the computed difference between the observed data and the synthesis (res)
## Not run: #Using TideCurve tc <- TideCurve(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2015/12/31", aetime = "23:30:00", ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") res_tc <- ResCurve(tc, tideObservation) #Using BuildTC and SynTC your_model <- BuildTC(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2016/12/31", aetime = "23:30:00", keep_data = TRUE) syn_tc <- SynTC(tmodel = your_model, ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") syn_tc$data_matrix <- your_model$data_matrix res_tc_n <- ResCurve(syn_tc, tideObservation) ## End(Not run)## Not run: #Using TideCurve tc <- TideCurve(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2015/12/31", aetime = "23:30:00", ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") res_tc <- ResCurve(tc, tideObservation) #Using BuildTC and SynTC your_model <- BuildTC(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2016/12/31", aetime = "23:30:00", keep_data = TRUE) syn_tc <- SynTC(tmodel = your_model, ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") syn_tc$data_matrix <- your_model$data_matrix res_tc_n <- ResCurve(syn_tc, tideObservation) ## End(Not run)
Synthesizes a tide curve; model built with BuildTC().
SynTC(tmodel = NULL, ssdate, sstime, sedate, setime, solar_syn = TRUE)SynTC(tmodel = NULL, ssdate, sstime, sedate, setime, solar_syn = TRUE)
tmodel |
The model you built with BuildTC(). Please see examples. |
ssdate |
Synthesis start date. This indicates the date you want your tide curve to start with. |
sstime |
Synthesis start time. The starting time for your tide table. |
sedate |
Synthesis end date. |
setime |
Synthesis end time. |
solar_syn |
Compute a solar synthesis? Default is TRUE. |
Returns a list with two elements, which are of class data.table and data.frame.
synthesis.lunar |
The lunar synthesis data as a data.table object in UTC. |
tide.curve |
The solar tide curve as a data.table or NULL object (time zone of the observations). |
## Not run: SynTC(tmodel = your_model, ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") ## End(Not run)## Not run: SynTC(tmodel = your_model, ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") ## End(Not run)
Takes a data frame as input with three columns (see example dataset) and returns a tide curve. Internally the analysis is carried out in lunar days. One mean lunar day lasts 1.0350501 mean solar days. Therefore the analysis time period should start one lunar day after the first observation and end one lunar day before the last observation.
TideCurve( dataInput, otz = 1, km = -1, mindt = 30, asdate, astime, aedate, aetime, ssdate, sstime, sedate, setime )TideCurve( dataInput, otz = 1, km = -1, mindt = 30, asdate, astime, aedate, aetime, ssdate, sstime, sedate, setime )
dataInput |
A data frame with the columns observation_date, observation_time and height. See attached data for correct formats. |
otz |
The time zone of the observations |
km |
The number of nodes between two consecutive mean moon transits. Shall be less or equal to: round(1440 [min] / time step [min]) Example: Time step 5 min: Use km = 288 or even smaller. Leave on default (km = -1) and supply mindt, when unsure. |
mindt |
Observation time step in [min]. Default is 30. |
asdate |
A string indication the date you want the analysis to start with. Format: "yyyy/mm/dd". |
astime |
A string indicating the time you want the analysis to start with. Format: "hh:mm:ss" |
aedate |
A string indication the date you want the analysis to end with. Format: "yyyy/mm/dd". |
aetime |
A string indicating the time you want the analysis to end with. Format: "hh:mm:ss" |
ssdate |
Synthesis start date. This indicates the date you want your tide curve to start with. Format: See above |
sstime |
Synthesis start time. The starting time for your tide table. Format: See above |
sedate |
Synthesis end date. Format: See above |
setime |
Synthesis end time. Format: See above |
Returns a list with elements of the analysis, fitting and the tidal curve for given data
synthesis.lunar |
The lunar synthesis data as a data.table object in UTC |
data.matrix |
The data needed for analysis |
tide.curve |
The solar tide curve as a data.table object (provided time zone) |
lm.coeff |
Coefficients for the km fitted linear models used in the synthesis as a list of 1-row matrices |
diff.analyse |
Time in days spanning the analysis |
Godin, Gabriel (1972) The Analysis of Tides. Toronto, 264pp
## Not run: TideCurve(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2015/12/31", aetime = "23:30:00", ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") ## End(Not run)## Not run: TideCurve(dataInput = tideObservation, asdate = "2015/12/06", astime = "00:00:00", aedate = "2015/12/31", aetime = "23:30:00", ssdate = "2015/12/17", sstime = "00:00:00", sedate = "2015/12/31", setime = "23:30:00") ## End(Not run)
A sample dataset containing observation date, time and height of high and low water
tideObservationtideObservation
A data frame with 10267 rows and 3 variables
date of observation, character value in "yyyy/mm/dd" format
time of observation, character value in "hh:mm:ss" format
observed value, numeric value