Title: ESM 266: Vegetation processes, biomass
1ESM 266 Vegetation processes, biomass
2Water balance for a landscape
Evapotranspiration (ET)
Precipitation (P)
slope
water table
Runoff (R Quickflow Delayed Flow)
(Quickflow leaves the landscape within a few
hours of rainfall. Delayed flow represents slow
drainage from soil water and ground water.)
3Conceptual model
Rnet
ET
- For some ?t (day, month) on a unit area of
land, the Continuity Equation is - U is the water content of the underground store
(soil or ground water) - Units are V/(AT) or depth/time (e.g., mm/mo, m/s)
Advection of sensible and latent heat
P
Quickflow
Soil
Recharge
Ground water
Delayed flow
4We need to evaluate evapotranspiration (ET)
- If we could evaluate ET we could use continuity
equation as an accounting procedure to calculate
runoff and soil moisture for any time period - We could account for the land-atmosphere
interaction, including any factor such as
vegetation type that might affect it - ET is extremely difficult to measure directly for
long periods, so we usually estimate it from the
energy balance (rate of expenditure of latent
heat) - Phase change to vapor requires an input of
2.5106 J/kg
5Basic principle
- Net radiation (Rnet) drives the sum of sensible
(H) and latent (L) heat exchange with the
atmosphere and heat flow into or out soil (G) - G is normally small
- Temperature, vapor pressure, and soil moisture
determine how Rnet is partitioned between H and L - i.e., the magnitude of the temperature gradient
vs. the vapor pressure gradient
6The energy balance equation (flux per unit area,
W m2)
- H sensible heat transfer ( is surface to
atmosphere) - L latent heat transfer in water evaporating or
condensing ( is evapotranspiration, is
condensation) - G heat conducted into soil
- S solar radiation
- a albedo water 0.06 conifer forest 0.09
Amazon broadleaf forest 0.12 grassland 0.20.4 - F? downward infrared radiation, depends on
temperature, water vapor, and clouds - Ts surface temperature
7Convert latent heat flux into mass flux of water
- Latent heat exchange per unit area converts a
volume of water (per unit area) to vapor - The energy required for this conversion is
- the volume of water per unit area (E)
- multiplied by the latent heat of vaporization
- energy required to convert a kg of water to vapor
- and by the density of water
- which converts the mass per unit area to a volume
per unit area
- L latent heat flux
- ET evapotranspiration rate (m s1)
- ?w density of water (1000 kg m3)
- ?v latent heat of vaporization (2.5?106 J kg1)
8End points (for day or month, G0)
9Partitioning between sensible and latent heat
exchange
using equations in the form of Sellers et al.
1997
- In equation form
- So were concerned with how
Controls latent heat exchange
Controls sensible heat exchange
10(variables)
11Problem 1, es (vapor pressure at leaf)
- Sellers et al. suggest
- The potential evapotranspiration is when ß1
- Common suggestion is
12Problem 2, resistance
- Sellers et al. (eq. 6) suggest both an
aerodynamic resistance ra and a surface
resistance rc
13MODIS Evapotranspiration Project (MOD16)
Kenlo Nishida Core science development, NTSG,
Univ. Montana / Univ. of Tsukuba Steve Running
and Rama Nemani Project directors, NTSG, Univ.
Montana Joe Glassy System development, Lupine
Logic Inc.
Comments? ? kenlo_at_ntsg.umt.edu
14Why estimate evapotranspiration by satellite?
- Score table of three approaches to ET
- --------------------------------------------------
--------------------------------------------------
---- Observation model satellite
RS-----------------------------------------------
--------------------------------------------------
------- - Time coverage Good! Good! Poor
- Spatial coverage Poor So-so. Good!
- Accuracy Good! So-so. So-so.
- Cost efficiency Poor So-so. So-so.
- --------------------------------------------------
--------------------------------------------------
---- - Particular advantage of satellite remote sensing
- Not influenced by water re-distribution (e.g.,
irrigation)- Strong at phenology
15Advantage of using MODIS on both Terra and Aqua
- Frequent global coverage (every day)
- Afternoon overpass for Aqua
- Dry conditions become clearer than morning
- Fine spatial resolution (0.5-1.0km)
- High-precision temperature measurement and
vegetation indices - Albedo and emissivity are also available
- Atmospheric information is available from other
sensors on Aqua - AIRS/AMSU/HSB give accurate atmospheric profile
16Outline of the Project
Final Product EF (Evaporation Fraction)
values - 8 day-period, 1km resolution,
globally - by Aqua (EOS-PM)/MODIS Requirements
for the algorithm - Stand alone It can operate
only with optical satellite sensor data -
Flexible It can ingest any other reliable data,
if available. - Simple It is simply constructed
to reduce computational load. - Scalable It can
give daily information of ET from instants
data. - Versatility It can operate regardless
of climate and biome. - Sensor Independence It
can co-operate with other sensors with same
logic.
17Definition of Evaporation Fraction (EF)
- Fractional value is representative for wetness
- Scalability of instantaneous observation to
longer period
From Crago, 1996, Scaling up in Hydrology using
Remote Sensing
18MODIS-ET Model Landscape
Actual landscape mixture of
forest, farm, grassland, road, etc.
Simplification
ETbare
ETveg
ET fveg ET veg (1 fveg) ET soil
Qveg
Qbare EF fveg ?? EFveg (1 fveg) ??
EFbare Q
Q
Fraction of bare soil 1.0 - fveg
Fraction of vegetation fveg
19How do we get it?
- Temperature on vegetation Tveg (S) - Incoming
solar radiation PAR (T)
- radiative transfer of atmosphere (T)
Qveg
Qbare EF fveg ??? EFveg (1 fveg)
??? EFbare Q
Q
We want this!
- Vegetation Index NDVI or EVI (S)
Note (S) . Derived from Satellite (T) .
Estimated theoretically
20Estimating EFveg (1) Concept
Central concept Evaporation from vegetation
(transpiration) is mostly controlled by stomata
opening (canopy resistance)
Assuming complementary relationship (ET PETPM
2PETPT PETPMPenmans PET PETPTPriestley-Tay
lors PET), we can get
Constant. 1.26
Derivative of saturated vapor pressure curve
(change with T)
a ? EFveg ??????????
? ? ( 1 rc / 2 ra)
Aerodynamic resistance
Psychrometric constant (slightly change with T)
Canopy resistance
21Estimating EFveg (2) Canopy Resistance Model
1 / rc f1(T) f2(VPD) f3(PAR) f4(?) / rcMIN
Ideally,
Solar radiation
Temperature
Soil water
Humidity
Actually,
Change of VI
1 / rc f1(T) f3(PAR) / rcMIN
22Estimating EFbare
satellite image
Ts
Tbare max
Warm Edge
Wind speed
Tbare
Window
Tbare max Tbare EFbare ??????? Tbare
max Tbare min
TvegTbare min
Air temperature
VI
VImax
VImin
VI
VI-Ts diagram (Nemani Running, 1989 1993)
Qbare0 ET Tbare
???????????? Ta 4esTa3 (1- CG) ?
Cp/ra bare
23Prototype
- NOAA/AVHRR 14-day composite (1km resolution)
- Window size 21km?21km
- Validation 13 sites of AmeriFlux
24(No Transcript)
25Only NDVI
Full algorithm
Test of simplified algorithm
Validation for each landcover type
26Prototype
Day233, 2000North USA (Tile H10-12V04)
27Remote sensing of biomass
- Carbon budget of a vegetation stand
- Approached through dynamic vegetation models
28Relationship of forest biomass to radar
backscatter
LeTuan et al., Climate Change, 2004
29Treuhaft et al., BioScience, 2004