3  Theme: Aquatic Ecosystems

What: Freshwater surface water temperature, oxygen, and chlorophyll-a.

Where: 7 lakes and 16 river/stream NEON sites.

When: Daily forecasts for at least 30-days in the future. New forecast submissions are accepted daily. The only requirement is that submissions are predictions of the future at the time the forecast is submitted.

Why: Freshwater surface water temperature, dissolved oxygen, and chlorophyll-a are critical for life in aquatic environments and can represent the health of the system

Who: Open to any individual or team that registers

How: REGISTER your team and submit forecast. If you registered for the Round 1 (2021) and are using the same team and method then you do not need to re-register.

The video below is an overview of the Aquatic Ecosystems Challenge that was recorded for the 2021 Early Career Annual Meeting

We held a Q&A session on May 21, 2021. You can find a recording from that session HERE.

3.1 Overview

In streams and rivers, forecasting water temperature can be meaningful for protecting aquatic communities while maintaining socio-economic benefits (Ouellet-Proulx et al. 2017). In lentic systems, successfully forecasting surface water temperatures can be important for fisheries and water utilities that need to manage the outflowing temperatures (Zhu et al. 2020). Recently, water temperature forecasts in lakes have been used to predict seasonal turnover when nutrients from the bottom can be mixed to the surface and impair the water quality.

Dissolved oxygen concentration is a critically important variable in limnology. Forecasts of dissolved oxygen in freshwaters is the first step to understanding other freshwater ecosystem processes. For example, oxygen serves as the gatekeeper to other biogeochemical reactions that occur in rivers and lakes. Preemptive forecasts of dissolved oxygen concentrations can anticipate periods of high or low oxygen availability, thereby providing insight into how the ecosystem may change at relatively short timescales.

chlorophyll-a is a metric of phytoplankton biomass. Phytoplankton biomass are the base of the aquatic food-web and an important indicator of water quality for managers.

3.2 Challenge

This design challenge asks teams to produce forecasts of mean daily surface water temperature and/or dissolved oxygen in 7 NEON lake and/or 16 NEON river/stream sites for 30-days in the future. Additionally, forecasts of chlorophyll-a are invited for the 7 lakes and 3 non-wadeable river NEON sites.

You can chose to submit to either the lakes, rivers or streams or all three. You can also chose to submit any of the three focal variables (temperature, oxygen, and chlorophyll).

Teams are asked to submit their 30-day forecasts of NEON surface mean daily surface water temperature, dissolved oxygen, and/or chlorophyll-a along with uncertainty estimates and metadata. NEON surface water temperature, dissolved oxygen, and chlorophyll-a collected prior to the current date will be provided and may be used to build and improve the forecast models. Other data can be used as long as teams provide access (minimum of URL, but ideally a script) to all teams in the challenge.

3.3 Data: Targets

The R script for generating the evaluation and training data (i.e., targets) can be found at: https://github.com/eco4cast/neon4cast-aquatics

The challenge uses the following NEON data products: - DP1.20264.001: Temperature at specific depth in surface water for lakes - DP1.20288.001: Water quality (includes oxygen and chlorophyll-a)
- DP1.20035.001{target =“_blank”}: Temperature in surface waters for streams

A file with previously released NEON data that has been processed into targets is provided below. The target script can be found here. The same processing will be applied to new data that are used for forecast evaluation.

Here is the format of the target file

readr::read_csv("https://data.ecoforecast.org/neon4cast-targets/aquatics/aquatics-targets.csv.gz") |> 
  na.omit()
# A tibble: 84,592 × 4
   datetime   site_id variable    observation
   <date>     <chr>   <chr>             <dbl>
 1 2016-08-12 ARIK    oxygen             3.40
 2 2016-08-13 ARIK    oxygen             4.16
 3 2016-08-14 ARIK    oxygen             4.07
 4 2016-08-15 ARIK    oxygen             3.91
 5 2016-08-16 ARIK    oxygen             3.86
 6 2016-08-17 ARIK    oxygen             4.35
 7 2016-08-17 ARIK    temperature       24.6 
 8 2016-08-18 ARIK    oxygen             4.35
 9 2016-08-18 ARIK    temperature       20.6 
10 2016-08-19 ARIK    oxygen             4.03
# … with 84,582 more rows

The target file has the following columns

  • datetime: date of observation
  • site_id: NEON site code
  • variable: variable (temperature, oxygen, chla)
  • observation: daily mean value

3.3.1 Surface Mean Daily Dissolved Oxygen Concentration

Definition

Dissolved oxygen (DO) is the concentration of oxygen dissolved in water. NEON’s 30-minute time resolution from deployed water quality sondes among the freshwater sites reports this concentration as mg L-1. We have adapted the available NEON DO data to output the mean daily DO concentration in mg L-1 from a water quality sonde(s) deployed in the top 1 m of the water column across all sites. Where multiple depths have observations within this depth range an average was taken. Common DO concentrations range between 0 and 12 mg L-1 and DO concentrations less than 2 mg L-1 are considered hypoxic.

Motivation

Dissolved oxygen concentration is a critically important variable in limnology. Forecasts of dissolved oxygen in freshwaters is the first step to understanding other freshwater ecosystem processes. For example, oxygen serves as the gatekeeper to other biogeochemical reactions that occur as well as determine the variety and health of aquatic organisms present in rivers and lakes. Preemptive forecasts of dissolved oxygen concentrations can anticipate periods of high or low oxygen availability, thereby providing insight into how the ecosystem may change at relatively short timescales.

3.3.2 Surface Mean Daily Water Temperature

Definition

Water temperature is the temperature of the water. NEON’s 30-minute time resolution from deployed water temperature sondes in the freshwater sites reports this in degrees Celsius (°C). We have adapted the available NEON water temperature data to output the mean daily water temperature in °C from temperature thermisters deployed 0-1m below the water surface at the lake sites and a water temperature sonde deployed in the stream sites. Where multiple depths have observations within this depth range an average was taken. Common water temperatures in lakes and streams range between 4 and 35 °C.

Motivation

In streams and rivers, forecasting water temperature can be meaningful for protecting aquatic communities while maintaining socio-economic benefits (Ouellet-Proulx et al. 2017). In lentic and lotic systems, successfully forecasting water temperatures can be important for management of fisheries and water utilities that rely on specific threshold temperatures (Zhu et al. 2020). Recently, lake temperature forecasts have been used to predict seasonal turnover, mixing bottom nutrients into the surface and impairing water quality.

3.3.3 Chlorophyll-a

Definition

chlorophyll-a (chla) is the concentration of chlorophyll-a in the water column, as measured using florescence. NEON’s 30-minute time resolution from deployed water quality sondes among the freshwater sites reports this concentration as mg L-1. We have adapted the available NEON chla data to output the mean daily chla concentration in mg L-1 from a water quality sonde deployed in the top 1 m of the water column at a lake sites and water quality sondes deployed in the non-wadeable river sites (BLWA, FLNT, TOMB). No ongoing measurements of chla are available in the wadeable streams and therefore cannot be forecasted in this challenge.

Motivation

Phytoplankton biomass are the base of the aquatic food-web and an important indicator of water quality for managers.

3.3.4 Focal sites

Information on the sites can be found here:

site_data <- readr::read_csv("https://raw.githubusercontent.com/eco4cast/neon4cast-targets/main/NEON_Field_Site_Metadata_20220412.csv") |> 
  dplyr::filter(aquatics == 1)
siteID site name waterbody type latitude longtitude NEON site URL
ARIK Arikaree River NEON Wadeable Stream 39.75821 -102.44715 https://www.neonscience.org/field-sites/arik
BARC Lake Barco NEON Lake 29.67598 -82.00841 https://www.neonscience.org/field-sites/barc
BIGC Upper Big Creek NEON Wadeable Stream 37.05972 -119.25755 https://www.neonscience.org/field-sites/bigc
BLDE Blacktail Deer Creek NEON Wadeable Stream 44.95011 -110.58715 https://www.neonscience.org/field-sites/blde
BLUE Blue River NEON Wadeable Stream 34.44422 -96.62420 https://www.neonscience.org/field-sites/blue
BLWA Black Warrior River NEON Non-wadeable River 32.54153 -87.79815 https://www.neonscience.org/field-sites/blwa
CARI Caribou Creek NEON Wadeable Stream 65.15322 -147.50397 https://www.neonscience.org/field-sites/cari
COMO Como Creek NEON Wadeable Stream 40.03496 -105.54416 https://www.neonscience.org/field-sites/como
CRAM Crampton Lake NEON Lake 46.20967 -89.47369 https://www.neonscience.org/field-sites/cram
CUPE Rio Cupeyes NEON Wadeable Stream 18.11352 -66.98676 https://www.neonscience.org/field-sites/cupe
FLNT Flint River NEON Non-wadeable River 31.18542 -84.43740 https://www.neonscience.org/field-sites/flnt
GUIL Rio Guilarte NEON Wadeable Stream 18.17406 -66.79868 https://www.neonscience.org/field-sites/guil
HOPB Lower Hop Brook NEON Wadeable Stream 42.47194 -72.32953 https://www.neonscience.org/field-sites/hopb
KING Kings Creek NEON Wadeable Stream 39.10506 -96.60383 https://www.neonscience.org/field-sites/king
LECO LeConte Creek NEON Wadeable Stream 35.69043 -83.50379 https://www.neonscience.org/field-sites/leco
LEWI Lewis Run NEON Wadeable Stream 39.09564 -77.98322 https://www.neonscience.org/field-sites/lewi
LIRO Little Rock Lake NEON Lake 45.99827 -89.70477 https://www.neonscience.org/field-sites/liro
MART Martha Creek NEON Wadeable Stream 45.79084 -121.93379 https://www.neonscience.org/field-sites/mart
MAYF Mayfield Creek NEON Wadeable Stream 32.96037 -87.40769 https://www.neonscience.org/field-sites/mayf
MCDI McDiffett Creek NEON Wadeable Stream 38.94586 -96.44302 https://www.neonscience.org/field-sites/mcdi
MCRA McRae Creek NEON Wadeable Stream 44.25960 -122.16555 https://www.neonscience.org/field-sites/mcra
OKSR Oksrukuyik Creek NEON Wadeable Stream 68.66975 -149.14302 https://www.neonscience.org/field-sites/oksr
POSE Posey Creek NEON Wadeable Stream 38.89431 -78.14726 https://www.neonscience.org/field-sites/pose
PRIN Pringle Creek NEON Wadeable Stream 33.37852 -97.78231 https://www.neonscience.org/field-sites/prin
PRLA Prairie Lake NEON Lake 47.15909 -99.11388 https://www.neonscience.org/field-sites/prla
PRPO Prairie Pothole NEON Lake 47.12984 -99.25315 https://www.neonscience.org/field-sites/prpo
REDB Red Butte Creek NEON Wadeable Stream 40.78393 -111.79789 https://www.neonscience.org/field-sites/redb
SUGG Lake Suggs NEON Lake 29.68778 -82.01775 https://www.neonscience.org/field-sites/sugg
SYCA Sycamore Creek NEON Wadeable Stream 33.75099 -111.50809 https://www.neonscience.org/field-sites/syca
TECR Teakettle Creek - Watershed 2 NEON Wadeable Stream 36.95593 -119.02736 https://www.neonscience.org/field-sites/tecr
TOMB Lower Tombigbee River NEON Non-wadeable River 31.85343 -88.15887 https://www.neonscience.org/field-sites/tomb
TOOK Toolik Lake NEON Lake 68.63069 -149.61064 https://www.neonscience.org/field-sites/took
WALK Walker Branch NEON Wadeable Stream 35.95738 -84.27925 https://www.neonscience.org/field-sites/walk
WLOU West St Louis Creek NEON Wadeable Stream 39.89137 -105.91540 https://www.neonscience.org/field-sites/wlou

3.4 Timeline

Forecasts for a minimum of 30 days can be submitted daily by 11:59 pm UTC. A minimum of 30 days in the future must be forecasted for each submission. For example, a forecast submitted on March 1 should be for at least March 1st – March 30th. New forecasts can be submitted daily as new weather forecasts and observations (e.g., new temperature and water quality data is released by NEON) become available. The key is that submissions are predictions of the future.

Since daily submissions are allowed and encouraged, the automation of forecast generation may be ideal. There are many ways to automate scripts that are written to download observations and meteorology drivers, generate forecasts, and submit forecasts. Two tools that many have used are cron jobs (see the R package cronR) that execute tasks at user specified times and github actions. See more at Chapter 12

Cron jobs work on unix and mac systems. An example of a script that executes a cron job using R can be found here.

3.5 Observed data latency

Through a collaboration with NEON, the new data will be available within 24-28 hrs of being collected for use in model training and forecast evaluation

3.6 Submissions

The required names for forecasted variables: oxygen, temperature, and chla.

The required time unit: date for daily forecast in YYYY-MM-DD

Instructions for submitting forecasts are found here: Chapter 2

3.7 Meterological inputs for modeling

Information about forecasted meteorology that is available for you to use when generating your forecasts can be found here: Chapter 9

3.8 Useful functions

Functions for validating, evaluating and submitting forecasts can be found here: Chapter 10

3.9 Null models

A climatology null model is automatically generated each day to served as a simple baseline model. This climatology null model forecasts that the nee or lee will be equal to the historical mean of that day of year.

Code for the climatology null model can be found here

3.10 FAQ

Answers to frequency asks questions can be found here: Chapter 12

3.11 Design Team

Freya Olsson, Virginia Tech
James Guinnip, Kansas State University
Sarah Burnet, University of Idaho
Ryan McClure, Virginia Tech
Chris Brown, National Oceanic and Atmospheric Administration
Cayelan Carey, Virginia Tech
Whitney Woelmer, Virginia Tech
Jake Zwart, United States Geological Survey

3.12 Partners

The challenge is hosted by the Ecological Forecasting Initiative (EFI; https://ecoforecast.org/) and its U.S. National Science Foundation sponsored Research Coordination Network (EFI-RCN; https://ecoforecast.org/rcn/).

Data used in the challenge are from the National Ecological Observatory Network (NEON): https://www.neonscience.org/.

Scientists from NOAA and USGS have been involved in the design of the challenge.

3.13 References

Ouellet-Proulx, S., St-Hilaire, A., and Bouchar, M.-A.. 2017. Water temperature ensemble forecasts: Implementation using the CEQUEAU model on two contrasted river systems. Water 9(7):457. https://doi.org/10.3390/w9070457

Zhu, S., Ptak, M., Yaseen, Z.M., Dai, J. and Sivakumar, B. 2020. Forecasting surface water temperature in lakes: a comparison of approaches. Journal of Hydrology 585, 124809. https://doi.org/10.1016/j.jhydrol.2020.124809