Transcript Bio-Climatic Model - World Conference on Climate Change

Possible changes of the brown trout habitat suitability in
the upper Po River basin due to global change.
A. Lombardi1, M. Verdecchia1,2, B. Tomassetti1, V. Colaiuda1 , A. Di Sabatino3
[1]{CETEMPS, Department of Physic and Chimestry, University of L’Aquila, L’Aquila, Italy}
[2]{Department of Physic and Chimestry, University of L’Aquila, L’Aquila, Italy}
[3]{Department of Life, Health and Environmental Sciences, University of L’Aquila, Italy}
Overview
Climatic Model
Grid Scale
RCM
General Circulation Model
(GCM) 100-200 km
CHyM
Regional Climatic
Model (RCM)
50-25 km
Downscaling
GCM
Bio-Climatic Model
Hydrological Model
< 1 km
RCM: RegCM Simulation
 RegCM has been driven
by the general circulation
model
ECHAM5
(Roeckner et al., 1996)
 RegCM has been forced
with A1B (Nakicenovic et
al.,
2000)
scenario
simulation, a MediumHigh scenario
 It simulated a period of
140 years, from 1960 to
2100.
 The simulation covered
Europe domain at 25 km
grid spacing.
 The simulation had a time
resolution of 3 hours
CHyM – CETEMPS Hydrological Model







Distributed grid-based deterministic hydrological model;
It includes an explicit parameterization of different physical processes
contributing to hydrological cycle, for each cell of the selected domain,
contribute of:
o
Runoff
o
Melting
Soil moisture
o
Evapotranspiration
o
Interception
storage
o
Infiltration
o
Rainfall
o
Return flow
are explicitly calculated.
It rebuilds precipitation field and drainage network using cellular
automata technique.
It runs in any geographical domain with any resolution up to DEM resolution, drainage network is extracted
by a native algorithm implemented in CHyM code;
Different sets of precipitation data can be assimilated and merged in a hierarchical way at each hourly time
step;
It runs in any Unix platforms;
It reads, in the current implementation, precipitation and temperature fields from:
o
o
o
o
RegCM model output,
MM5 model output,
WRF model output,
ERA-interim reanalysis
CHyM Applications
Tacina River Basin
25 Sept 2009
00:00 – 12:00 AM
Prediction
monitoring
Effects and
of Climate
Change
on hydrological cycle
Climatic Change on upper Po basin
According to many climate change projections
produced with global and regional climate models,
the Italian peninsula will experience pronounced
changes in temperature and precipitation (Coppola
and Giorgi, 2010 and Giorgi, 2006).
The hydro-climatic simulations show
• a lengthening of the dry season, which would
increase the water stress for the area.
• the largest signal is in winter, when the discharge
increases everywhere in the basin, and
particularly in high elevation areas, as a result of
increased magnitude and a liquid fraction of
precipitation (Coppola and Giorgi, 2010).
The Po River water resources are indeed vulnerable
to future climate changes, which should be taken into
account in the development of suitable adaptation
options in terms of water management for the basin
(Coppola et al. , 2014).
From GCM to Bio-Climatic Model
Could expected changes in hydrometeorological condition on the
Po basin affect a specific specie?
GCM
RCM
CHyM
The first aim of this preliminary study is
represented by development of detailed
numerical approach to estimate how the
fitness of our target specie is linked to hydroclimatic condition.
Bio-Climatic Model
Bio-Climatic Model
The target specie used is adult brown trout.
Brown trout is among the world’s top invasive species, displacing
other fish species through both competition and predation (Lowe
et al., 2000; McIntosh et al., 2011)
Its presence is based on three parameters (Jorde, 1996;. Capra et
al, 1995; Heggenes et al, 1996a, b; Peviani et al., 1996; Bartsch
et al., 1996; Boudreau et al., 1996; Bovee, 1986):
•
Water Velocity
•
Water depth
•
Water temperature.
The
Bio-Climatic
Model
uses
hydrological model outputs and USGS HSC
(Habitat Suitability Criteria) curve interpolation
to associate to each grid point of the domain a
value beetween 0 and 1.
𝑛
𝑃 𝑥 =
𝑛
𝑓 𝑎𝑖
𝑖=1
𝑗≠𝑖
𝑥 − 𝑎𝑗
𝑎𝑖 − 𝑎𝑗
How are these parameters calculated?
WVELOCITY =
S1/2 R2/3
n(μ)
R = β + γDAδ
W𝐃𝐄𝐏𝐓𝐇 =
Q
α
∗H
𝐓𝐰𝐚𝐭𝐞𝐫 (𝐭) = A + B ∗ Ta (t)
•
•
•
•
S is longitudinal bed slope of the flow element ,
n the Manning’s roughness coefficient
R is the hydraulic radius. It is a linear function of the drained area DA.
β, γ and 𝛿 are empirical constants to tune with during the calibration. The
exponent 𝛿 is usually very close to 1
• V is costant in the time; it depends only on the geomorphology
• Q is flow discharge estimated by the model
• 𝛼 is water velocity
• H is an empirical coefficcient. It is equal to 0.3
• 𝐴 = 4.90; 𝐵 = 0.60
• 𝑆 = 𝑛𝑖=1(𝑦𝑖 − 𝑓 𝑥𝑖 )2
Water Temperature
•
Hourly long time series of Air
temperature (10 years X 10 sensor
stations) .
•
Daily long time series of water
temperature (almost 3 years X 10
sensor stations).
𝑇𝑤𝑎𝑡𝑒𝑟 (𝑡) = 𝐴 + 𝐵 ∗ 𝑇𝑎 (𝑡)
𝑛
(𝑦𝑖 − 𝑓 𝑥𝑖 )2
𝑆=
𝑖=1
𝐴 = 4.90; 𝐵 = 0.60
Hourly obserbed air temperature
Daily observed air temperature
Daily observed water temperature
Daily estimated water temperature
Overall Suitability Index
𝑂𝑆𝐼(𝑖,𝑗) = 𝑇𝑆𝐼(𝑖,𝑗) ∗ 𝐷𝑆𝐼
𝑖,𝑗
∗ 𝑉𝑆𝐼(𝑖,𝑗)
salmo trutta, Linnaeus, 1758
salmo [trutta] trutta (Zerunian, 2004)
Areal distribution in Italy
salmo [trutta] trutta(Ruffo and Stoch, 2006)
Presence signs in Italy
salmo trutta, Linnaeus, 1758 in upper river Po basin
salmo [trutta] trutta(Ruffo and Stoch, 2006)
Overall Suitability Index Trend
More than 𝟐 𝑿 𝟏𝟎𝟏𝟏
data were analyzed.
•
Brown trout is among the world’s top invasive
species, displacing other fish species through
both competition and predation (Lowe et al.,
2000; McIntosh et al., 2011) .
•
The sexually mature fishes move to upstream of
rivers and smaller tributaries.
•
It is both migratory and territorial. Usually fry,
after only one year of life, moves down to the
valley, where, as adults, it prefers to remain,
seeking out those areas with higher food
availability, directly connected to a sufficient
flow discharge, a mild water temperature and
minimum flow velocity that guarantee good
oxygenation of the water and less pollution.
•
Renata E Hari, David M Livingstone, Rosi Siber,
PATRICIA BURKHARDT-HOLM, and Herbert Guettinger.
Consequences of climatic change for water
temperature and brown trout populations in alpine
rivers and streams. Global Change Biology, 12(1):10–
26, 2006.
•
William J Matthews and Earl G Zimmerman. Potential
effects of global warming on native fishes of the
southern great plains and the southwest. Fisheries,
15(6):26–32, 1990.
Conclusion
Climatic simulation models predict an increase in temperature and extreme events occurrence. These changes are expected
to lead a sensible modification of the hydrological cycle with significant impacts on the ecological integrity of aquatic
ecosystems. Changes in temperature regime and instream habitat/microhabitat characteristics will also affect the natural
distribution of many aquatic species.
To this aim we carried out a simulation based on a chain of models to predict the distribution of the brown trout in the
upper Po River basin (North Italy). A 140-years long simulation, carried out with a Regional climate model, is used to
force a hydrological model simulating the hydrological cycle. The results of hydrological simulation, in particular
variations in temperature and discharge regimes, are then used to evidence the areas where the target species is expected to
occur.
The results show how the complex proposed approach can reproduce, with a good confidence, the habitat suitability of the
brown trout. The projection for future years indicates a shift of the distribution toward locations of the upper part of the
basin, with a sensible decrease of the areas where the brown trout can survive, reproduce and grow.
It appears strategic to predict the effects of global change on freshwater biodiversity and species distribution in order to
propose adequate measures aimed at mitigating the impacts of climate modification on natural systems.
The work also focuses on the potential application of the proposed approach to evaluate the effects of climate changes on
more complex ecological systems.
CHyM (Cetemps Hydrological Model) – Parametrization of physical
processes contributing to hydrological cycle.
Based on the kinematic wave approximation (Lighthill and Whitam, 1955) of the shallow water
wave the equations used by CHYM model to simulate the surface routing overland and for channel
flow are the continuity and momentum conservation equations:
𝜕𝐴 𝜕𝑄
+
= 𝑞𝑐
𝜕𝑡
𝜕𝑥
Q = 𝛼𝐴𝑚
Where A is the flow cross-sectional area, Q is the flow rate of water discharge, 𝒒𝒄 is the rate of lateral
inflow per unit of length, t is the time, x is the coordinate along the river path, 𝜶 is the kinematic wave
parameter, and m the kinematic wave exponent usually assumed = 1.
The kinematic wave parameter has the dimension of a speed and it can be written as:
1
𝛼=
2
𝑆 2𝑅 3
𝑛
where β, γ and 𝛿 are empirical constants to
tune with during the calibration. The exponent
d is usually very close to 1.
S is longitudinal bed slope of the flow element , n
the Manning’s roughness coefficient while R is
the hydraulic radius that can be written as a
linear function of the drained area DA as:
𝑅 = β + γ𝐷𝐴𝛿