Primary productivity in the Arctic: a model developed at Takuvik

# Primary productivity in the Arctic: a model developed at Takuvik
## 🛰️ Presented by Philippe Massicotte
### In collaboration with (no specific order): Maxime Benoît-Gagné, Simon Bélanger, Emmanuel Devred, Atsushi Mastuoka, Marcel Babin, Marie-Hélène Forget, Jean-Thomas Baillargeon, Eric Rehm, Mathieu Ardyna, (and probably many others I forgot)
### 2020-02-11–2020-02-12 (Ottawa, Canada)

---

# A brief presentation of the primary production model developed at Takuvik

---

# Modeling primary production

Daily primary production `\((mgC~m^{-2}~d^{-1})\)` can be *easily* calculated as follows:

`\begin{equation}
\text{PP} = P^B_{\text{max}}\int_{t~=~0}^{24h}\int_{z~=0}^{z_\text{max}}\text{Chl a}(z)[1-e^{\frac{-\text{PUR}(z,t)}{E_k\text{PUR}(z)}}]\,\mathrm{d}t\mathrm{d}z
\end{equation}`

**Important to have proper values/estimates of Chl *a*.**

But why do we need another primary production model 🤔

---

# Estimating Chl *a* in the water column

- Ocean color (OC) satellite sensors can only detect surface Chl *a*.

- Can not detect subsurface chlorophyll maxima (SCMs) that can significantly contribute to areal primary production in the Arctic Ocean (*Hill et al., 2005; Weston et al., 2005; Martin et al., 2010*).

- **Hence, SCMs are generally omitted from primary production estimates derived from remote sensing.**

---

# Subsurface chlorophyll maxima (SCMs)

---

# Subsurface chlorophyll maxima (SCMs)

---

# Parameterization of vertical chl *a* in the Arctic Ocean

The study of Ardyna et al. 2013 tackled the problem of characterizing the SCMs in the Arctic.

> In the Arctic Ocean, deep subsurface chlorophyll maxima (SCMs) that significantly contribute to primary production (PP) are often observed. These are neither detected by ocean color sensors **nor accounted for in the primary production models applied to the Arctic Ocean**.
>
> — *Ardyna et al., 2013*

---

# Taking into account the vertical variations in Chl *a*.

---

# How to model Chl *a* vertical profile?

*Ardyna et al. 2013* proposed to parameterize the vertical Chl *a* profiles using a Gaussian-based equation:

`\begin{equation}
c(z) = C_b - sz + C_{\text{max}}e^{-(\frac{z - Z_{\text{max}}}{\Delta z})^2}
\end{equation}`

How to estimate the parameters `\(C_b, s,  C_{\text{max}},  Z_{\text{max}}\)`?

**This is the main novelty of the Takuvik's primary production model.**

---

# How to model Chl *a* vertical profile?

Parameter values determine the shape of the SCMs.

---

# Bloom state, region, depth and surface Chl *a*

Flowchart developed by Ardyna et al. (2013).

---

# Model specifications

---

# Model specifications

- Climate Change Initiative (CCI)
    - Merged MERIS, Aqua-MODIS, SeaWiFS and VIIRS
    - **Surface Chl *a***, Rrs

- MODIS
    - Atmospheric products (cloud cover, cloud optical thickness, ozone) to calculate downwelling radiance `\((E_d)\)`
    - Lookup table from Simon Bélanger

---

# Model specifications

**Important:** 
  
  - All parameter inputs can be modified
  
  - Opens the door for data with better spatial/temporal coverage
  
---

# A pan-Arctic model

- Pan-Arctic (45N+) divided into 27 regions (based on WWF definition).

- Daily between 2003-2018 (15 years of continuous data).

---

# A pan-Arctic model

---

# Wrapping-up

- Data has been produced for the 2003-2015 period.
    - Available on request, but plan to put it online when we find a way to share such a huge database.

- Easy to re-run with *better* inputs.
    - See Juan Li's presentation entitled *retrieval of chlorophyll-a concentration in the Arctic Ocean*.