10-wind.Rmd

# Wind {.tabset}

```{r wind_options}
knitr::opts_chunk$set(cache = FALSE)
```

## Time Series

Something to look for: ws/wd lock due to rime on the anemometer or vane, data coming online after long periods (multiple hours) of being offline only to have the first hour read high, then all subsequent hours much lower.

### Wind Speed

```{r wspd_sclr-hourly}
wspd_sclr_plots<-if(
  
  (data %>% 
       dplyr::filter(PARAMETER %in% toupper("wspd_sclr") & !is.na(RAW_VALUE)) %>%
       nrow(.))==0
  
) {htmltools::tags$p("There is no data for this parameter.")} else {
  
  plotly_fcn(data,
                        all_data,

            parameter=toupper("wspd_sclr"))
  
  
}

htmltools::tagList(wspd_sclr_plots)
```

### Wind Direction

```{r wdir_vect-hourly}
wdir_vect_plots<-if(
  
  (data %>% 
       dplyr::filter(PARAMETER %in% toupper("wdir_vect")& !is.na(RAW_VALUE)) %>%
       nrow(.))==0
  
) {htmltools::tags$p("There is no data for this parameter.")} else {
  
  plotly_fcn(data,
                        all_data,

            parameter=toupper("wdir_vect"))
  
  
}

htmltools::tagList(wdir_vect_plots)
```

## Histograms

### Wind Speed (Scalar)

```{r ws-histogram}

suppressWarnings(
  htmltools::tagList(
  
  purrr::map(data |>
             dplyr::pull(STATION_NAME) |>
             unique(x=_),
  
  function(station)
    
  
    {
   
    plot_data<-data |> 
             dplyr::filter(STATION_NAME %in% station &
                             PARAMETER %in% toupper("wspd_sclr"))
    
    if(nrow(plot_data)!=plot_data |> 
      dplyr::summarise(nas=sum(is.na(RAW_VALUE))) |>
      dplyr::pull(nas)){
    
    p <- ggplot2::ggplot(data = plot_data,
                  ggplot2::aes(x = RAW_VALUE,
                      fill = INSTRUMENT,
                      color = INSTRUMENT)) +
        
        ggplot2::geom_histogram(alpha = 0.25,
                       binwidth = 1) +
        
        ggplot2::labs(
          title = paste("Histogram:",
                        station),
          y = "Count",
          x = "Hourly Wind Speed (m/s)"
        ) +
        ggplot2::theme_bw() +
        ggplot2::theme(legend.position = "none",
              legend.title = ggplot2::element_blank()) +
        ggplot2::scale_fill_brewer(palette = "Set1") +
        ggplot2::facet_wrap(~ INSTRUMENT, scales = "free")
      
      plotly::ggplotly(p)  
    
    }  else{htmltools::tags$p("There is no data for this parameter.")}#end if and else
      
    } #end plot function
  
  )# end map
  
) #end tagList

) #end suppressWarnings
```

### Wind Direction (Vector)

```{r wd-histogram}

suppressWarnings(
  htmltools::tagList(
  
  purrr::map(data |>
             dplyr::pull(STATION_NAME) |>
             unique(x=_),
  
  function(station)
    
  
    {
   
    plot_data<-data |> 
             dplyr::filter(STATION_NAME %in% station &
                             PARAMETER %in% toupper("wdir_vect"))
    
    if(nrow(plot_data)!=plot_data |> 
      dplyr::summarise(nas=sum(is.na(RAW_VALUE))) |>
      dplyr::pull(nas)){
    
    p <- ggplot2::ggplot(data = plot_data,
                  ggplot2::aes(x = RAW_VALUE,
                      fill = INSTRUMENT,
                      color = INSTRUMENT)) +
        
        ggplot2::geom_histogram(alpha = 0.25,
                       binwidth = 1) +
        
        ggplot2::labs(
          title = paste("Histogram:",
                        station),
          y = "Count",
          x = "Hourly Wind Direction (degrees)"
        ) +
        ggplot2::theme_bw() +
        ggplot2::theme(legend.position = "none",
              legend.title = ggplot2::element_blank()) +
        ggplot2::scale_fill_brewer(palette = "Set1") +
        ggplot2::facet_wrap(~ INSTRUMENT, scales = "free")
      
      plotly::ggplotly(p)  
    
    }  else{htmltools::tags$p("There is no data for this parameter.")}#end if and else
      
    } #end plot function
  
  )# end map
  
) #end tagList

) #end suppressWarnings
```

## Wind Roses

Wind roses for `r params$year` and 5 previous years (when available). 

Things to look for when validating wind data:

+ Annual windrose
+ Windrose from previous years
+ Percent calm winds for the past 5-8 years
+ Do we have any pictures of the station?
+ Does the windrose make sense to you relative to the local topography?
+ RH - does RH get to 100% or does it max out at 97%? Has this changed recently?
+ Can I see any obvious breakpoints in a strip chart where something changes/

```{r wind-cache}
knitr::opts_chunk$set(cache = FALSE)
```

```{r wind-data-prep-qaqc}
# select wind data, get it in openair format

wind_data <- # contains X previous years as per _01.import-data.html
  dplyr::bind_rows(
    data |> # data from import-data chunk in index.Rmd
      
      dplyr::filter(PARAMETER %in% c("WSPD_SCLR", "WDIR_VECT")),
    prev_yrs_wind
  )
```

```{r duplicates,eval=FALSE,include=FALSE}
# # # # # FIND DUPLICATES
 duplicates<-wind_data |>
    dplyr::group_by(DATE_PST, STATION_NAME, PARAMETER) |>
    dplyr::summarise(n = dplyr::n(), .groups = "drop") |>
    dplyr::filter(n > 1L) |>
   dplyr::left_join(x=_,
                    wind_data,
                    by=c("DATE_PST","STATION_NAME","PARAMETER"))

utils::View(duplicates)

# save duplicates as csv
 write.csv(duplicates,"duplicates.csv")

 
 # # # EXPLORATORY DATA ANALYSIS # # #
 
 #look at the first 6 rows of each set of duplicates (by PARAMETER)
 duplicates |>
   dplyr::group_by(PARAMETER) |>
   dplyr::slice(1:6) |> utils::View()

  # are any groups all NA's? makes it easy to remove
 duplicates |>
   # dplyr::filter(PARAMETER=="WSPD_SCLR") |>
   dplyr::group_by(PARAMETER,INSTRUMENT) |>
   dplyr::summarise(`# NA's`=sum(is.na(RAW_VALUE)),
                    `# Obs`=sum(!is.na(RAW_VALUE)),
                    `# Hours`=dplyr::n())


# when did the duplicates start and end?
duplicates |>
  dplyr::group_by(STATION_NAME,PARAMETER,INSTRUMENT) |>
  dplyr::summarise(start=min(DATE_PST),
                   end=max(DATE_PST))

# to look at the whole data set (sorted by PARAMETER and DATE_PST)
duplicates |>
  dplyr::arrange(PARAMETER,DATE_PST) |> utils::View()


 
duplicates |>
  dplyr::filter(PARAMETER=="WDIR_VECT") |> utils::View()
 # odd rows - shoudl be all NA's
      dplyr::slice(seq(1,nrow(.),2)) |> utils::View()

# preview wind_data
wind_data |>
  dplyr::filter(INSTRUMENT=="UNSPECIFIED" &
                  lubridate::year(DATE_PST)==2021) |> utils::View()
 
 # # # END DUPLICATES
  
```

```{r wind-data-openair}
if( all(nrow(wind_data)!=0, # at least one row of data
        any(unique(wind_data$PARAMETER) %in% "WSPD_SCLR"),
        any(unique(wind_data$PARAMETER) %in% "WDIR_VECT")
    )
  ){
  
  wind_data <- wind_data |>
    
    dplyr::select(date = DATE_PST,
                  STATION_NAME,
                  # INSTRUMENT,
                  PARAMETER,
                  RAW_VALUE) |>
    
    tidyr::pivot_wider(data=_,
                       names_from = PARAMETER,
                       values_from = RAW_VALUE) |>
    
    dplyr::rename(ws = WSPD_SCLR,
                  wd = WDIR_VECT) |>
    
    dplyr::mutate(ws = ifelse(is.na(wd), NA_real_, ws),
                  wd = ifelse(is.na(ws), NA_real_, wd))
}


# openair::summaryPlot(wind_data)
```

```{r wind-rose, fig.show='hide'}
if(all(
  # at least a row of data
  nrow(wind_data) != 0, 
  # some data in the year being validated
  length(unique(lubridate::year(wind_data$date)) %in% params$year)!=0,
  # a column called ws
  any(names(wind_data) %in% "ws"),
  # a column called wd
  any(names(wind_data) %in% "wd")
         
       )
) {
 
  
  station<-unique(wind_data$STATION_NAME)

   
plot_list<-lapply(unique(lubridate::year(wind_data$date)) |> sort(x=_),
                  
# plot_list<-purrr::walk(unique(lubridate::year(wind_data$date)) |> sort(x=_),
    
    function(year){
      
      # TESTING
      
      # year<-2024
      # 
      # wind_dataBackup<-wind_data
      # 
      # wind_data<-wind_dataBackup
      
      # END TESTING
      
      
      wind_data %<>%
        dplyr::filter(lubridate::year(date) %in% year)
      
      #number of calm hours
      ncalm <- wind_data |>
        dplyr::filter(ws < 0.5) |>
        dplyr::summarise(calms = n())
      
      #total number of valid ws hours
      ntotal <- wind_data |>
        dplyr::filter(!is.na(ws)) |>
        dplyr::summarise(total = n())
      
      #percentage calm (ws<0.5 m/s)
      pcalm <- round(100 * ncalm / ntotal,
                     digits = 2)
      
      #filter out calm wind_data
      roseData <- wind_data |>
        dplyr::filter(ws >= 0.5)
      
      if(nrow(wind_data)==sum(is.na(wind_data$ws)) | nrow(roseData)==0){
        
        #make an empty plot when there is no data for a year
        ggplot2::ggplot(data=tibble::tibble(
          y=NA_real_,
          x=NA_real_),
          ggplot2::aes(x=x,y=y
        ))+
          ggplot2::annotate("text",x=0,y=0,label="There is no valid paired wind data for this year.")+
          ggplot2::labs(x="",
                        y="",
                        title=year)+
          ggplot2::theme_bw()+
          ggplot2::theme(axis.text.x = ggplot2::element_blank(), 
                axis.ticks.x = ggplot2::element_blank(), 
                axis.text.y = ggplot2::element_blank(), 
                axis.ticks.y = ggplot2::element_blank()) +
          ggplot2::coord_polar()
        
        
      } else {
      
      
      
      #windRose
      p<-windRose(
        roseData,
        annotate = FALSE,
        breaks = c(0.5, 1.5, 3.3, 5.5, 7.9, 10.7, 13.8, 17.1),
        #Beaufort scale with 0.5 as lowest cut point.
        sub = paste("Calms (=<0.5m/s)=", pcalm, "%"),
        key.position = "right",
        main = stringr::str_c(year),
        angle = 360 / 16,
        #16 spokes
        cols = "jet",
        paddle = FALSE
      ) 
      
      p$plot
      
      } # end else
 
    } #end function(station)
     ) # map
 
} else{
  
  plot_list<-htmltools::tags$p("This station either doesn't measure wind, has no wind data for the year being validated, or it is missing one of WSPD_SCLR or WDIR_VECT")
  
}


```

```{r wind-rose-fig-dims}
#define width and height of wind rose figures based on length of plot_list

fig_height<-dplyr::case_when(
  length(plot_list) %in% 1:2 ~ 4,
  length(plot_list) %in% 3:4 ~ 8,
  length(plot_list) %in% 5:6 ~ 12,
  length(plot_list) %in% 7:8 ~ 16
  
)

```

```{r wind-rose-output,fig.height=fig_height}
if(is.null(plot_list$name)){
  gridExtra::grid.arrange(grobs=plot_list,
                        ncol=dplyr::if_else(length(plot_list)==1,1,2),
                        top=unique(wind_data$STATION_NAME))
}else{
  plot_list 
}
```