Langtjern

Lake Langtjern (60°37’N; 9°73’E) is a small and shallow (lake surface 0.227 km2, mean depth 2 m) humic, acid-sensitive, oligotrophic lake located in central Norway. Langtjern has a small catchment area (4.8 km2, ca. 510-750 m.a.s.l) that is dominated by unproductive pine forest, wetlands and bedrock. Yearly precipitation is 750 mm/yr. Langtjern generally experiences long winters and a stable snowpack.

Data

These data were used to model lake carbon processes with the GOTM-FABM family of models. The formats used by GOTM-FABM are described in https://github.com/fabm-model/fabm/wiki/GOTM

Meteorological data

Wind speed, air temperature, relative humidity and precipitation were taken from NIVA's Langtjern long-term ecological monitoring site.

Air temperature and relative humidity are measured using a Campbell Scientific 215. A Young Model 05103-5 measures wind speed and direction. Precipitation is measured with a Geonor T-200B. No local measurements of mean sea level pressure and cloud cover were available but were interpolated with data from the Norwegian Meteorological Institute network of stations.

Filename: langjern-weather.dat

Meterological forcings

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Wind speed north-south (m/s)
• Wind speed east-west (m/s)
• Mean sea level pressure (millibars)
• Air temperature (^oC)
• Relative Humidity (%)
• Cloud Cover (%)
• Precipitation (mm)

Carbon data

Color Dissolved Organic Matter (CDOM) is measured with a TriOS Microflu-CDOM (ex 370nm/em 460nm) (D, small sensor) and UV-absorption with a TriOS ProPS (190-360 nm) (D, large sensor) with a 1 cm light path. Both have mechanical wipers to remove biofilm.

The raw CDOM data has been converted to DOC data with a two-step process: 1) first the CDOM data has been corrected for temperature quenching according to the temperature correction equation recommended by Ryder et al. (2012), then 2) a Gaussian process regression was used to convert the CDOM data into DOC data.

1. The CDOM at a reference temperature (CDOMref) is given by:
   
   CDOMref=CDOM×(1+m/(Tmeas×m+C) ×(Tref−Tmeas))
   
   where CDOM is the raw CDOM signal, Tref is the reference temperature (e.g., 20°C), Tmeas is the measured water temperature, and m and C are the slope and intercept, respectively, of any given regression equation of temperature versus CDOM fluorescence determined in the lab. Since this regression hasn’t been determined for Langtjern, m and C have been optimized to provide the best correlation between CDOMref and weekly DOC data. A simple optimization genetic algorithm written in MATLAB® has been used for this purpose.

2. A Gaussian process regression can be used as a non-linear multivariate interpolation tool. In this case, high-frequency DOC data has been expressed as a function of up to 8 predictors including CDOMref, as well as water level and water temperature integrated (or not) over various timescales. In a Gaussian process regression, the coefficient (or vector) for each predictor is variable across time. Different predictors were used for the inflow and outflow.

Inflow to the lake

Filename: langtjern_carbon_inflow.dat

Total carbon inflow to the lake.

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Carbon concentration (mg/m³)

Outflow from the lake

Filename: langtjern_carbon_inflow.dat

Total carbon outflows from the lake

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Carbon concentration (mg/m³)

Oxygen data

Oxygen saturation is measured (O₂) with an Aanderaa Optode 4175 at depths of 1m and 8m. Oxygen saturation is transformed into a concentration using the water temperature (T) according the formula:

O₂ / 100.0 * ( (14.59 - 0.3955 T + 0.0072 T²- 0.0000619 T³) / 31.9988 )

Filename: langtjern_O2.obs

Oxygen concentrations at 1 and 8m.

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Depth (m)
• Concentration (mmol/m³)

Discharge and water temperature data

Gauge height (gh) is measured with a pressure transducer which is then transformed into discharge according to the following formulas.

• inlet: 2.391(gh-0.345)^2.5
• outlet: 3.2136(gh-0.315)^2.453

The discharge thus calculated is scaled according to the area it drains compared to the total area draining to the lake.

Water temperature is measured using a thermocouple.

Filename: langtjern-inlet-q.dat

Discharge into the lake and water temperature.

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Discharge (m³/s)
• Water temperature (^oC)

Filename: langtjern-outlet-q.dat

Discharge from the lake and water temperature.

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Discharge (m³/s)
• Water temperature (^oC)

Solar radiation

Global radiation is measured with an Apogee SP-212 sensor. This can be used instead of the cloud cover by the models.

Filename: langtjern-radiation.data

Total incoming solar radiation.

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Irradiance (W/m²)

Lake water temperature

The lake Langtjern buoy measures water temperature using a thermocouple at different depths.

Filename: langtjern_lake_temperature.dat Water temperature in the lake.

Columns:

• Date: YYYY-MM-DD HH:MM:SS
• Depth (m)
• Water temperature (^oC)