NEM calculation
The change in oxygen concentration in the water reflects the balance between photosynthetic production, respiratory consumption, and physical exchange at the water-atmosphere interface. By measuring dissolved oxygen (DO) concentrations at high frequency with in situ data loggers, we can calculate gross primary production (GPP), ecosystem respiration (ER), and net ecosystem metabolism (NEM), as detailed by Staehr et al. (2010).
Load and process the raw data file
First, we need to obtain the necessary parameters from the raw data
downloaded from the data loggers. We will remove unnecessary rows and
columns and extract the dissolved oxygen concentrations and water
temperature, along with the corresponding date and time recorded by the
logger using process_hobo(). This function will also
eliminate any NA values in the dataframe. The no_hobo input
was necessary for further analysis. Currently, it supports raw data
downloaded from the Onset HOBO Dissolved Oxygen Data Logger
(U26-001).
hobo_data_path <- system.file("extdata", "ex_hobo.csv", package = "aelab")
do <- process_hobo(hobo_data_path, no_hobo = "code_for_logger")
do[c(1:3), ]## date_time do temp no_hobo
## 1 2024-04-08 12:00:00 7.69 25.86 code_for_logger
## 2 2024-04-08 12:30:00 7.67 26.02 code_for_logger
## 3 2024-04-08 13:00:00 7.65 26.20 code_for_loggerThe air pressure and wind speed at the sampling site should be
obtained either manually or from a nearby weather station. The
process_weather function tidies the CSV file downloaded
from Taiwan’s weather station. In addition to extracting the columns
with air pressure and wind speed values, it also duplicates the hourly
data to match the half-hourly recorded frequency of the DO
concentrations.
weather_data_path <- system.file("extdata", "ex_weather.csv", package = "aelab")
weather <- process_weather(weather_data_path, date = "2024-04-10", zone = "zone_A")
weather[c(1:5), ]## pressure_hpa wind_ms date_time zone
## 1 1015.2 1.4 2024-04-10 00:30:00 zone_A
## 2 1015.2 1.4 2024-04-10 01:00:00 zone_A
## 3 1015.0 1.6 2024-04-10 01:30:00 zone_A
## 4 1015.0 1.6 2024-04-10 02:00:00 zone_A
## 5 1014.9 1.4 2024-04-10 02:30:00 zone_AIn addition to the previously mentioned parameters, we also needed
the water depth, salinity, start and end date and time of the data
logger deployment, and the sunrise and sunset times during that period.
These parameters can be entered in Excel and then imported into R using
process_info() to ensure that all necessary parameters are
included. The column zone corresponds to
process_weather(), and column no_hobo
corresponds to process_hobo().
info_data_path <- system.file("extdata", "info.xlsx", package = "aelab")
info <- process_info(info_data_path)
info## # A tibble: 1 × 9
## zone site no_hobo depth_m salinity start_date_time end_date_time
## <chr> <chr> <chr> <dbl> <dbl> <dttm> <dttm>
## 1 zone_A site_A code_f… 1 10 2024-04-10 00:00:00 2024-04-10 23:59:00
## # ℹ 2 more variables: sunrise <dttm>, sunset <dttm>Finally, we can merge the dataframe containing air pressure and wind
speed data from process_weather() with the processed data
downloaded from the data loggers. Next, we will combine this with the
other necessary parameters using the zone and
no_hobo columns. We can use the plot_hobo()
function to inspect the changes in DO concentrations.
data <- merge(weather, do, by = "date_time")
merged_df <- data %>%
inner_join(info, by = c("zone", "no_hobo"))
merged_df[c(1:3), ]## date_time pressure_hpa wind_ms zone do temp no_hobo
## 1 2024-04-10 00:30:00 1015.2 1.4 zone_A 8.28 25.10 code_for_logger
## 2 2024-04-10 01:00:00 1015.2 1.4 zone_A 8.07 25.04 code_for_logger
## 3 2024-04-10 01:30:00 1015.0 1.6 zone_A 7.80 24.98 code_for_logger
## site depth_m salinity start_date_time end_date_time
## 1 site_A 1 10 2024-04-10 2024-04-10 23:59:00
## 2 site_A 1 10 2024-04-10 2024-04-10 23:59:00
## 3 site_A 1 10 2024-04-10 2024-04-10 23:59:00
## sunrise sunset
## 1 1899-12-31 05:45:00 1899-12-31 18:15:00
## 2 1899-12-31 05:45:00 1899-12-31 18:15:00
## 3 1899-12-31 05:45:00 1899-12-31 18:15:00