Components of near-surface energy balance derived from satellite soundings – Part 1 : Net available energy

Introduction Conclusions References


Introduction
An important manifestation of climate change is widespread alterations to the composition of the energy balance at the Earth's surface.Given the importance of being able to predict the consequences of climate change, both measurement and modelling of the components of surface energy balance attract significant attention from a broad range of related scientific disciplines.Two such disciplines are hydrology and meteorology, which share a common interest in resolving the balance between sensible, H, and latent, λE , heat fluxes over a broad range of spatial and temporal scales.
Net available energy, Φ, is a core variable used to predict the magnitude of H and Figures

Back Close
Full λE given it is defined as the sum of these two fluxes (Wright et al., 1992), The utility of this definition arises from being able to also specify Φ as the difference between the net broadband radiation, R N , and the rate of heat accumulation, G, below the plain across which R N is specified, Given R N is routinely measured using net radiometers this affords an opportunity to specify Φ and hence either H or λE .For example, in modelling studies λE is invariably specified as a function of Φ using now ubiquitous functions such as those of Penman (1948) for open water or Monteith (1965) for land surfaces.Despite being the rate of change of a heat stock, in terrestrial environments G is often interpreted as the "ground heat flux" and attempts to measure this using heat flux plates are commonplace (Mayocchi and Bristow, 1995;Sauer and Horton, 2005).These measurements prove somewhat less reliable than R N due to greater spatial heterogeneity in ground heat uptake (Gao et al., 1998) allied to the fact that significant heat capacity resides in other elements of the land surface.As a result, G proves problematic in surface energy balance studies and is either ignored (Foken et al., 2006) despite being significant under a broad range of conditions (Santanello and Friedl, 2003), or treated somewhat superficially (Choudhury, 1987).
The arrival of satellite retrievals for many of the components of R N has opened up opportunities to develop large scale estimates of this variable and hence λE (Batra et al., 2006;Mu et al., 2007).For example, retrievals for the components of R N have been available through the International Satellite Cloud Climatology Project (ISCCP) (Pinker and Laszlo, 1992); the Earth Radiation Budget Experiment (ERBE); Clouds and Earth's Radiant Energy System (CERES) on board of the NASA Earth Observing System (EOS) and Tropical Rainfall Measuring Mission (TRMM) satellites (Wielicki et al., 1998).Several studies have reported the estimation of R N using a combination Introduction

Conclusions References
Tables Figures

Back Close
Full of MODIS (MOderate Resolution Imaging Spectroradiometer) atmospheric and land products over USA, China, and India (Bisht et al., 2005;Cai et al., 2007;Mallick et al., 2009) or NOAA-14 (National Oceanic and Atmospheric Administration) data over the Tibetan Plateau (Ma et al., 2002).
Unfortunately, in the absence of direct observations of G at spatial scales and coverage of satellite R N , retrievals for Φ have had to rely on the parameterisation of G using surface temperature, albedo and vegetation index information (Bastiaanssen et al., 1998;Batra et al., 2006) or by assigning some fixed proportion of R N (Choudhury, 1987;Humes et al., 1994) in satellite-based surface energy balance models (Mecikalski et al., 1999).But studies by Kustas et al. (1993) and recently Hsieh et al. (2009) demonstrated that G is, by definition, a highly dynamic quantity, and that the ratio G/R N can range anywhere from 0.05 to 0.50 depending on the time of day, soil moisture and thermal properties, and vegetation density.Therefore, methods that are able to provide defensible estimates of G in conjunction with R N in order to give Φ directly from satellite data without having to rely unduly on any offline calibration would clearly be of great benefit to this area.In this paper we present a method for retrieving R N and Φ based on exploiting both satellite radiance data and day-night surface temperature difference.Taking advantage of the extensive network of terrestrial eddy covariance tower sites (Baldocchi et al., 2001) which record direct measurements of R N , H and λE , we use these to derive independent non-radiative estimates of Φ in order to critically evaluate our satellite estimates of this quantity.

Net radiation
The approach for estimating R N uses the Atmospheric Infrared Sounder (AIRS) radiation products, although we have also made use of the MODIS surface reflectance Introduction

Conclusions References
Tables Figures

Back Close
Full following balance between net shortwave (R NS ) and longwave (R NL ) radiation.
where α is the surface albedo and R L↓ and R L↑ are the downwelling and upwelling thermal radiative fluxes (all fluxes specified in W m −2 ).Our chosen reference level for R N is the near surface given this corresponds to the flux-based tower estimates we used in the evaluation.Therefore, surface R S was estimated from its top-of-atmosphere clear sky counterpart R S0 and AIRS cloud cover fraction (f ) following Hildebrandt et al. (2007), where τ A is the clear sky transmissivity of the atmosphere which we assume is 0.75 (Cano et al., 1986;Thornton and Running, 1999;Hildebrandt et al., 2007).The terrestrial surface albedo was generated using the MODIS Aqua-Terra surface reflectances r i following Liang et al. (1999), where r i are the mid-point reflectances within the 0.62-0.67; 0.841-0.876; 0.459-0.479; 0.545-0.565; 1.230-1.250; 1.628-1.653;and 2.105-2.155µm wavelength bands and p i are the weightings for each wavelength bands taken as p i =(0.3973; 0.2382; 0.3489; −0.2655; 0.1604; −0.0138; 0.0682) (Liang et al., 1999).The albedo of ocean varies according to the cosine of solar zenith angle (Jin et al., 2004).In the present case, a constant albedo of 0.04 was assumed for oceans given the satellite radiances are nadir.
Clearly, many of the longwave components of the radiative balance are very closely related to the raw IR radiances being measured by AIRS.Given these are not in the public domain we have attempted to recover them as follows, although in future we Introduction

Conclusions References
Tables Figures

Back Close
Full would anticipate using the raw IR radiances more directly if possible.R NL was calculated as, where σ is the Stefan-Boltzmann constant (5.67×10 −8 W m −2 K −4 ), T C is the columnar air temperature, and ε C and ε S are the column and surface emissivity, respectively.
Among the different schemes for calculating ε C we have used that proposed by Prata (1996) given this appears to be the most reliable (Bisht et al., 2005).This scheme uses AIRS total precipitable water (ξ) information to estimate ε C as, The columnar air temperature T C in Eq. ( 6) is taken as the average of the 2 m and 1000 hPa pressure level AIRS temperatures in an attempt to reflect a weighting toward the lower troposphere when specifying R L↓ .T S and ε S are taken directly from the AIRS skin temperature and surface emissivity products.

Surface heat capacity, ground heat flux and net available energy
The definition of G stems from consideration of the non-steady state surface energy balance, where c is the aggregate surface system heat capacity.The AIRS sounder platform samples twice daily at 01:30 and 13:30 LT.Despite being somewhat coarse, taking a discrete time, backward difference approximation of Eq. ( 8) with a sample interval ∆t=12 h equivalent to that of the AIRS pass gives,

Conclusions References
Tables Figures

Back Close
Full If we assume that the system is approximately in equilibrium over a 24 h cycle, and that at 01:30 LT Φ≈0 (Tamai et al., 1998), then this gives the following simultaneous equations: which can be solved analytically to derive b 1 and b 2 and hence Φ and c for each grid cell in the AIRS global array.
Clearly Eq. ( 10) is a coarse approximation of Eq. ( 8) and hence potentially suffers from a number of deficiencies.Firstly, diurnal symmetry in ∆T s is only appropriate when one considers weekly or monthly average behaviour, and that there are no additional heat losses or gains to/from stores beyond the domain defined by the single heat capacity c.In this study we will be looking at monthly average behaviour because AIRS only gives partial global coverage on the daily timescale due to both cloud effects and the non-overlapping swath width of the sensor.Interactions with additional long term heat stores is an issue in systems such as the oceans where there can be a persistent heat loss/gains to/from deeper water over timescales of weeks to months, although relative to the diurnal fluctuation of stored surface heat this tends to be small (Stramma et al., 1986).Secondly, Φ can be either positive or negative at 01:30 LT, although it tends to be only a fraction of its 13:30 LT value due to the supply of energy being relatively small at night compared to the day.This may be less true for areas of land in the height of winter with cloud full days (as we get in Western Europe) and over the sea where significant daytime heat accumulation could in part be re-released as night time latent and sensible heat.Finally, all the terms in Eq. ( 8

Satellite datasets
In the present study two different data sources were used for the estimation of R N and Φ, AIRS and MODIS.were obtained in hierarchical data format (HDF4) along with their latitude-longitude projection.
We have used the MODIS Aqua atmospheric product datasets (MYD08 D3) (http: //modis-atmos.gsfc.nasa.gov/index.html) at 1 • ×1 • spatial resolution for extracting the solar zenith angle field.For generating the surface albedo fields we used narrowband surface reflectances from combined MODIS Terra-Aqua 16 day data (MCD43C4) products acquired from the MODIS data archive (http://ladsweb.nascom.nasa.gov/data/search.html).The native spatial resolution of the MODIS datasets is 0.05 • .Therefore, all the narrowband surface reflectances were first resized into 1 • by 1 • to make them compatible with the AIRS spatial resolution and then the broadband surface albedo was generated from these data.Introduction

Conclusions References
Tables Figures

Back Close
Full

Evaluation of R N and Φ
To evaluate the satellite values of R N and Φ we have made use of the extensive FLUXNET terrestrial tower network (Baldocchi et al., 2001).Clearly, there is a scale conflict here with the satellite retrievals being 1 • whilst the tower observations are for scales of the order of 1 km or less.The tower R N are from the broadband net radiometer sensors located on each tower.In the absence of reliable measures of G at the tower scale and in order to derive genuinely independent measures of Φ against which to evaluate the satellite data, we have taken the tower net available energy as the sum of the measured sensible and latent heat flux i.e.Eq. ( 1).Thereby we have assumed that the eddy covariance flux measurements are able to close the energy balance (i.e. R N −G=λE +H), the implications of which will be discussed below.We have chosen 30 sites covering a broad range of geographical locations selected from 7 land cover types including; evergreen broadleaf forest (EBF), mixed forest (MF), evergreen needle forest (ENF), deciduous broadleaf forest (DBF), savanna (SAV), grassland (GRA) and cropland (CRO).A comprehensive list of the site characteristics are provided in Table 1.Each tower evaluation dataset is comprised of the 13:30 LT samples of R N , H and λE which correspond with the satellite overpass.Again, the evaluation is based on pooling these data into weighted monthly average values.For the evaluation we have elected to compare all 12 months of data for 2003 given this had the best overlap between the FLUXNET and AIRS databases.

Results
The locations of the 30 terrestrial evaluation sites are marked in Fig. 1. retrieval and hence are rejected although these will be less evident in the monthly or annual average data.
Figure 2a shows the global distribution of R N which generally decreases with latitude as expected.R N also decreases over land due to the generally higher albedo resulting in reduced absorbtion of the net shortwave radiation (Giambelluca et al., 1997) or relatively higher surface temperature increasing the net longwave component, especially over the drier regions (Liang et al., 1998).As a result the magnitude of R N was around 200-300 W m −2 over the dry desert regions whereas the oceanic values of R N were 450-700 W m −2 .
Figure 2b shows the global distribution of c.The oceanic values of 4-8 MJ m are equivalent to 1-2 m of sea water, which appears reasonable on the daily time step to which they relate (Stramma et al. 1986).These oceanic values are somewhat noisy due to the small day-night temperature differences observed for the oceans giving a relatively poor signal to noise ratio.However, behind this noise the pattern of oceanic c appears relatively uniform as one might expect.Over land c varies between 0.05-0.5 MJ m −2 K −1 with wetter tropical and high latitude areas showing significantly higher values than the drier, less vegetated areas as expected.The soil equivalent depth of this heat capacity is approximately 0.01 m, which again appears reasonable for a daily time step (Li and Islam, 1999), although in heavily vegetated areas c is obviously comprised of a more complex aggregation.
Figure 2c shows the global distribution of G.These are the 13:30 LT values, hence them being net positive as an annual average.Between 20 • North-South G is approximately 10 to 20% of R N , and this rises to more than 40% above 50% North-South (Hsieh et al., 2009).Given this opposes the pattern of R N one would conclude either some deficiencies in the way G is specified here or that R N partitions into latent heat far more effectively than surface heating in these warm wet environments (Liu et al., 2005).
Again, terrestrial values are lower than their oceanic equivalents due in the main to the lower heat capacity as well as reduced R N as discussed above.This also highlights the role of the vegetation layer in preventing ground heating (Baker and Baker, 2002).Introduction

Conclusions References
Tables Figures

Back Close
Full The Sahara appears particularly prominent in this scene with high rates of midday heat accumulation which appears to be associated with a combination of moderate net radiation and relatively high heat capacity.The heterogeneity in this region appears to be related to the pattern of bare darker rock.
Figure 2d shows the global distribution of Φ which follows a similar pattern to R N as expected, although the pattern of G shown in Fig. 2c dictates that the North-South gradients in Φ are somewhat stronger than those of R N .Before discussing these results we consider their evaluation.Figure 3a shows the pooled evaluation of R N which produced an overall correlation of r=0 Figure 3b shows the evaluation for Φ which produced pooled statistics of r=0.87 (±0.03),Φ(satellite)=0 90(±0.03)Φ(tower)- 2.43(±8.19)and RMSD of 72 W m −2 .The biome specific statistics for Φ are also given in Table 2 showing correlations ranging Figure 4 shows a sample of monthly time series for Φ for both the satellite and the towers.The sites were selected to represent the biome classes considered here and also ones for which complete annual data sets for 2003 were available.These results show the satellite estimates generally track the trends in the tower data and hence the pooled statistics are not masking the within site variability.Again, the site-wise comparative statistics for these data are given in Table 2. Introduction

Conclusions References
Tables Figures

Back Close
Full There have been very few attempts to retrieve satellite estimates of Φ and compare these with ground truth data, although the statistics from our attempt appear to parallel results of Stisen et al. (2008) who studied a single site in the Senegal River basin using 5 km spatial resolution MSG (Meteosat Second Generation) geostationary satellite data and obtained a correlation of 0.51 and an RMSD of 14 to 17% of the mean in comparison to the surface measurements.
As seen in Fig. 3a  cloudy sky conditions.As a result, any bias in the shortwave is not as a result of biased sampling when comprising the monthly average.Besides, this would tend to lead to R S (satellite)>R S (tower).
There are many possible reasons R S (satellite)<R S (tower).Firstly, one has to question the accuracy of Eq. ( 4).Assuming the measure of cloud cover fraction is accurate, the assumed clear sky transmissivity of 0.75 into Eq.( 4) may be too low, although we have avoided any attempt to recalibrate this based on the tower data given it is not clear this is the principle source of the observed bias.Beyond Eq. ( 4), the diffuse fraction of R S (tower) can become enriched by surface reflected solar radiation, particularly in undulating terrain (Dubayah and Loechel, 1997).It would also be tempting to implicate nonlinear scaling effects of surface albedo in R S (satellite)<R S (tower) (Salomon et al., 2006) because, despite appearing as a linear component in Eq. ( 3), surface albedo interacts nonlinearly with other dynamic surface variables such as surface wetness and land surface temperature (Ryu et al., 2008) or the leaf area index (Hammerle et al., 2008).However, given the observed bias in R N appears to be explained through the downwelling shortwave one would conclude that here this effect is of secondary importance.
To probe the specification of R N further we investigated the individual longwave radiation components in relations to measures of theses fluxes available for a limited subset (14) of tower sites.From Fig. 3d,e it appears that there is quite good agreement between the satellite and tower data for both R L↓ and R L↑ and that any mismatch is insufficient to explain the discrepancy in R N .This is somewhat surprising for two reasons.Firstly, unlike the shortwave component, R L (tower) is all sky whilst R L (satellite) is only from clear sky conditions where IR retrieval is possible.As a result one would anticipate very significant differences in the monthly average values of the longwave components.However, it is difficult to predict the effect of this biased sampling on R NL (satellite) given cloud interacts with both R L↓ and R L↑ in complex ways.Secondly, one would anticipate significant scaling effects from the T 4 nonlinearity in Eq. ( 6) which can result in a disproportionate contribution of warmer elements within the landscape Introduction

Conclusions References
Tables Figures

Back Close
Full to both R L↓ and R L↑ (Kustas and Norman, 2000;Lakshmi and Zehrfuhs, 2002).The fact that these affects are not seen to any significant degree could point to compensating errors in the analysis but does not distract from the central message that of the importance of the bias in the shortwave when accounting for R N (satellite)<R N (tower).Figure 3b and Table 2 show that Φ(satellite)≈0.90Φ(tower)suggesting a slight compensation for the underspecification of R S through the underspecification of G in the satellite data.However, this evaluation assumes the energy balance to be closed in the tower data (i.e.R N −G=λE +H), which typically is not the case, λE +H often falling short of R N −G by 20% (Wilson et al., 2002).Because the causes of this energy imbalance remain controversial (see Foken, 2008 for a recent review), it is difficult to estimate how much the tower values of λE +H are actually biased low and hence the extent to which this bias affects our evaluation.By way of illustration, if, in the worst case, the entire energy imbalance was to be attributed exclusively to λE +H (i.e.R N −G are quantified correctly), then the true midday λE +H could be some 20% greater (Wilson et al., 2002).As a result, the present bias seen in Table 2 would change to Φ(satellite)≈0.72Φ(tower)again implicating R S as the main source of bias in the satellite retrievals for both R N and Φ.

Conclusions
We have only evaluated the satellite retrievals using data from terrestrial sites, and clearly it would be worthwhile repeating this for the ocean retrievals if possible.We have held back on this evaluation here because of the lack of an extensive network of instantaneous latent and sensible heat flux or radiative flux data over the oceans, although we note that the SEAFLUX project within the Global Energy and Water Experiment (GEWEX) initiative should give rise to such a database in the near future.
From the terrestrial evaluation we would argue that the methodology employed here shows some promise for specifying both R N and Φ, although the results suggests the need for improvements particularly in the specification of R S .More detailed studies Introduction

Conclusions References
Tables Figures

Back Close
Full evaluating the representativeness of each tower site footprint in relation to the 1 • scale scene within which it is situated could prove useful in this regard as would methods for cloud-proofing the satellite retrievals under persistent cloudy sky conditions.
Having derived and provisionally evaluated Φ for terrestrial systems this is available to schemes for estimating latent and sensible heat using satellite data.Given we have resorted to the minimal amount of calibration in deriving Φ it would appear sensible if a similar philosophy were adopted in developing satellite-based schemes for these important fluxes.With the availability of high spatial and temporal resolution geostationary imager -sounder suite from GOES and future GIFTS (Geosynchronous Interferometric Fourier Transform Spectrometer) as well as INSAT (Indian National Satellite)-3-D our present approach could be extended to derive Φ at every half hourly time scale.
In addition to opportunities in specifying large scale surface heat and vapour fluxes, the heat capacity estimates made here clearly carry information on variations in terrestrial surface moisture storage and we envisage that studies to develop this concept further could prove fruitful, particularly because of the emergence of satellite microwave data against which the results could be compared.Introduction

Conclusions References
Tables Figures

Back Close
Full   Full     For details of the site characteristics see Table 1.For the comparative statistics see Table 2.The solid line is the pooled linear regression given in Table 2.For details of the site characteristics see Table 1.For the comparative statistics see Table 2.The solid line is the pooled linear regression given in Table 2.
Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | where ∆T s is the day-night surface temperature change, b 1 =∆t/c and b 2 =−Φ(t)∆t/c.
) are highly dynamic and yet are treated as constant or varying linearly over the 12 h sample interval.It is difficult to predict what the consequences of this are given it depends on the pattern radiative forcing through the day.Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Figure 2 shows annual average, global satellite scenes for 13:30 LT R N , c, G and Φ for 2003.Missing data in the images are mainly due to missing data in the AIRS soundings at high latitudes or over the mountain belts where it is difficult to profile air temperature and relative humidity reliably.In addition, persistent cloudy conditions also prevent reliable Introduction Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | from 0.70 (EBF) to 0.95 (ENF), RMSD ranging between 62 (GRA and SAV) to 88 (EBF) W m −2 and regression coefficients ranging between 0.66(±0.08)to 1.01(±0.05)and −65.25(±27.07) to 108.71(±32.10)for the gain and offset, respectively.

For
R N the statistics relating the satellite and tower data are comparable with the results of:Bisht et al. (2005) who obtained 74 W m −2 RMSD when evaluating MODIS Terra geophysical land products over the Southern Great Plains of the US;Jacobs et al. (2004) who obtained a 12.2% error when evaluating GOES (Geostationary Operational Environmental Satellite) geostationary data over Southern Florida; andCai et al. (2007) who obtained 13.7% error when evaluating MODIS Terra-Aqua data over China.Stackhouse et al. (2000) evaluated the International Satellite Cloud Climatology Project (ISCCP) data to have errors in the range 10 to 15 W m −2 in monthly average shortwave and longwave radiative fluxes.When these errors are compounded in the derivation of R N and compared with tower data, an RMSD of the order of 98 W m −2 appears reasonable.
there is a systematic underestimation of R N relative to the tower values which exceeds the typical accuracy of net radiometer measurements of 20 W m −2 quoted by Foken (2008).We examined this underestimation in more detail by, where possible, evaluating the three individual radiation components of R N (R S , R L↓ and R L↑ ).All tower sites provided measurements of R S .Figure 3c shows this is systematically underestimated at the satellite scale with R S (satellite)= 0.70(±0.02)RS (tower)+68(±12.24)which could account for the mismatch of R N (satellite)≈0.75R N (tower).Before attempting to account for the various reasons for this underestimation it is important to realise that, unlike the IR components, the shortwave components are all-sky retrievals i.e. like the tower data they do not omit Introduction Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Ibrom, A Discussion Paper | Discussion Paper | Discussion Paper | Hollinger, D. Y., Goltz, S. M., Davidson, E. A., Lee, J. T., Tu, K., and Valentine, H. T.: Seasonal patterns and environmental control of carbon dioxide and water vapour exchange in an ecotonal boreal forest, Glob.Change Biol., 5(8), 891-902, 1999.Humes, K. S., Kustas, W. P., and Moran, M. S.: Use of remote sensing and reference site measurements to estimate instantaneous surface energy balance components over a Discussion Paper | Discussion Paper | Discussion Paper | Stramma, L., Cornillon, P., Weller, R. A., Price, J. F., and Briscoe, M. G.: Large diurnal sea surface temperature variability: satellite and in situ measurements, J. Geophys.Res., 16, 827-837, 1986.Tamai, K., Abe, T., Araki, M., and Ito, H.: Radiation budget, soil heat flux and latent heat flux at the forest floor in warm, temperate mixed forest, Hydrol.Process., 12, 2105-2114, 1998Discussion Paper | Discussion Paper | Discussion Paper |

Figure 1 .
Figure 1.The distribution of the 30 eddy covariance tower sites used for evaluating RN and Φ. 630

Fig. 1 .
Fig. 1.The distribution of the 30 eddy covariance tower sites used for evaluating R N and Φ.

Figure 3 .
Figure 3.Comparison of satellite and tower monthly average 13:30 hour a. RN. and b.Φ.For details of the site characteristics see Table1.For the comparative statistics see Table2.The solid line is the pooled linear regression given inTable 2. Comparison of satellite and tower monthly average 13:30 hour c.RS , d. RL↓ and e. RL↑ for a selection of sites for

Fig. 3 .
Figure 3.Comparison of satellite and tower monthly average 13:30 hour a. RN. and b.Φ.For details of the site characteristics see Table1.For the comparative statistics see Table2.The solid line is the pooled linear regression given inTable 2. Comparison of satellite and tower monthly average 13:30 hour c.RS , d. RL↓ and e. RL↑ for a selection of sites for 27 Fig. 3. Comparison of satellite and tower monthly average 13:30 LT (a) R N and (b) Φ.For details of the site characteristics seeTable1.For the comparative statistics see Table 2.The solid line is the pooled linear regression given in Table 2. Comparison of satellite and tower monthly average13:30 LT (c) R S , (d) R L↓ and (e) R L↑ for a selection of sites for which tower data for R S (360 data points), R L↓ (159 data points) and R L↑ (159 data points) were available.The linear fit (solid line) between the two sources of R S is, R S (AIRS)=0.70(±0.02)RS (tower)+67.68(±12.24);r=0.84(±0.03).The linear fit (solid line) between the two sources of R L↓ is, R L↓ (AIRS)= 1.03(±0.03)RL↓ (tower)−36.91(±10.05);r=0.95(±0.03).The linear fit (solid line) between the two sources of R L↑ is, R L↑ (AIRS)=0 91(±0.02)R L↑ (tower)+20.43(±8.77);r=0.96(±0.02).The dashed lines are 1:1 in all cases.(+ EBF; x MF; • GRA; * CRO; ∇ ENF; ♦ DBF; SAV).
The AIRS sounder was carried by the NASA Aqua satellite, which was launched into a sun-synchronous low Earth orbit on 4 May 2002 as part of the NASA Earth Observing System.It gives near global coverage twice daily at 01:30 a.m./p.m. from an altitude of 705 km.AIRS produces 3-dimensional maps of air and surface temperature, water vapor and cloud properties.Level 3 standard monthly day-night data products of air temperature and relative humidity profiles, cloud cover fraction, surface emissivity, near-surface air temperature and surface-skin temperature and columnar total precipitable water at 1 • ×1• spatial resolution were obtained for 2003 from the online data archive of AIRS, distributed through NASA Mirador data holdings (http://mirador.gsfc.nasa.gov/).The monthly products are simply the arithmetic mean, weighted by counts, of the daily data of each grid box.The multi-day merged products have been used here because the IR retrievals are not cloud proof and the multi-day product gave decent spatial cover in light of missing cloudy sky data.The data products

Table 1 .
Eddy covariance sites used for the evaluation of the satellite derived R N and Φ.