- Earth’s Changing Surface
- NSTA Science Store :: Earth's Changing Surface: Sculpting the Landscape :: Science Object
- Main navigation
- GEOG 272 Earth's Changing Surface (3 credits)
Retrieved September 23, from www. The results suggest that recent changes in global vegetation have had The researchers also created a new method by which scientists Below are relevant articles that may interest you. ScienceDaily shares links with scholarly publications in the TrendMD network and earns revenue from third-party advertisers, where indicated. Living Well. View all the latest top news in the environmental sciences, or browse the topics below:. Embeds 0 No embeds. No notes for slide. Earth's Changing Surface 1.
One reason it changes is because of Continental Drift. Continental Drift is the theory of how the continents move over the surface of Earth. Over the years, the land broke apart and became separate continents. It is made up of many plates that float on the surface of Earth. As the plates move around, they cause changes to the land.
When plates scrape and slide past each other, an earthquake occurs. Some of the minerals in rocks dissolve in water. Water can also get in cracks in the rock and freeze. When it freezes, it expands and breaks the rock apart. Wind that contains sand wears rock away and creates tiny rock pieces. The roots of plants can spread inside cracks in a rock. This loosens the rock and pushes pieces of rock apart. Bits of rock soon break off. For example, the Grand Canyon was carved out by the Colorado River, but it took about six million years. Sediment is tiny pieces of rock, soil, and sand.
Earth’s Changing Surface
Erosion carries weathered materials away from a place. Glaciers cut away the land and carry pebbles, rocks, and even huge boulders to a new spot. For example, Half Dome in Yosemite was cut in half by a glacier. Deposition happens when the speed of water decreases or slows down and the sediment drops to the bottom of the water. Earthquakes happen along faults or on the edges of different plates. Faults are breaks in a plate where pieces of rock move. In the ocean, an earthquake can cause a tsunami.
A volcano is a mountain formed by lava and ash. Hot, melted rock called magma comes out of the Earth. When magma reaches the surface of Earth, it is called lava. Lava cools and hardens. When this happens, rock is formed. This changes the shape of Earth. The input surface property variables need to have a common spatial resolution 0. For all variables, the median value for each month is calculated from the years to to generate the 5-year climatology while retaining the seasonal cycle. The source and pre-processing of each individual variable is as follows.
Day-time and night-time LST are the radiant temperature of a surface measured in the day or at night, respectively. These times are close to those at which the minimum and maximum temperatures are expected. Surface upwelling longwave radiation is the outgoing infrared radiation emitted by the surface. A large part of this energy is absorbed by the atmosphere and later re-emitted towards the Earth by clouds and greenhouse gases or outwards to space. We use day-time and night-time LST from the MYD11C3 product to estimate the mean surface temperature T over the entire day span using a simple average.
Albedo is defined as the proportion of the incident light or radiation that is reflected by a surface. The NASA MCD43C3 albedo product provides 8-daily estimates of both directional hemispherical albedo black-sky albedo and bihemispherical albedo white-sky albedo based on multidate multispectral MODIS cloud-free observations collected over a day moving window and a semi-empirical kernel-driven bidirectional reflectance model These white-sky and black-sky albedos correspond to theoretical situations in which incident radiation is either completely diffuse or completely direct.
To obtain an estimate of real conditions without information on the fraction of diffuse radiation, we took the mean of both values. To have estimates at monthly temporal resolution, we selected only those in which the day periods correspond best with the 15th of each month. In this case, we are interested in the terrestrial component associated with plant transpiration.
The product is not entirely observation-driven as it requires some specific parametrization per biome, but which is not spatially explicit. To identify the biophysical signal due to changes in vegetation cover we establish a relationship between vegetation cover fractions and the surface variables over a local moving window.
As a result of this, the direct biophysical effects of vegetation change considered here are local.
This is valid both for the spatial extent of the cover change, which assumes at most a change of a complete fine resolution pixel 0. The moving window size is 5 by 5 pixels at 0. There is a problem, however, if the compositional predictor data set X is used directly in the analysis. Analysis of any given subset of compositional components can lead to very different patterns, results and conclusions Geometrically, all points defined by the compositions must fall in a simplex because their compositions sum to one.
For a three part composition, this simplex is a triangular plane i. While the matrix has 3 columns, there are only at most 2 dimensions. A transformation of X is needed to reduce appropriately the dimensionality of this matrix for subsequent use in the regression. The transformation we apply to reduce the dimensionality of X involves a singular value decomposition SVD. This procedure is very close to a principal component analysis PCA. The first step consists of centring all the columns of the predictor matrix X of vegetation fractions by removing the column means.
We then apply the SVD:. Squared values of D indicate how much variance is explained by each orthogonal dimension. In doing so, we reduce the dimensionality appropriately as described above, as well as remove what may be additionally redundant dimensions that can occur locally if, for instance, the only points in which 2 classes are represented have exactly the same values. To avoid having problems when there is too little or no information e.
The final appropriately transformed predictor matrix of reduced dimension Z is then obtained by:. The resulting predictor matrix Z can now be regressed onto the local biophysical variable y. Because the compositional predictor matrix X has been transformed to matrix Z , regression coefficients identified in the regression of Z onto y do not immediately provide information about the association between the various vegetation cover fractions and the surface property variables.ipdwew0030atl2.public.registeredsite.com/73864-best-track.php
NSTA Science Store :: Earth's Changing Surface: Sculpting the Landscape :: Science Object
This is the y associated with that vegetation type. Since we wish to do this for all compositional components of interest, we actually define a matrix P with as many rows as these compositional components that we wish to predict. P is centred on the same column means as above M , specific to each local analysis , and then multiplied by the correct number of transposed right hand singular vectors V z , again, specific to each local analysis.
The expected change in variable y associated with a transition from one vegetation type to another at the central pixel of the local window is then the difference between the y p predicted for each pure vegetation type:. We consider uncertainty in terms of standard deviations, and thus, according to error propagation, the uncertainty for the difference due to the transition from A to B can be determined from:. This covariance term is important as the uncertainties of the individually predicted z values are not independent given that they derive from the same regression model.
The variances and covariances of all vegetation types can be obtained from the covariance matrix, which in turn is calculated as:. The whole procedure described above variable transformation, regression and uncertainty estimation is applied globally over 5 by 5 moving windows for the 3 biophysical variables for each of the 12 months of the year at 0.
Symmetric transitions yield identical results e. The resulting maps only provide information for the pixels in which all 25 pixels in the moving window had information. The method relies on there existing co-occurrences of vegetation classes within the local window. Furthermore, the statistical methods that are applied to these sets of points are more likely to provide reliable results when there are large and balanced presences of both vegetation classes of interest. Another masking operation is required to remove areas where high topographical variability exists within the local window.
Topographical relief generally translates into climatic gradients, which would compromise the space-for-time approach. For more information on this masking step, readers are again directed to the data descriptor paper The maps resulting from the local space for time analysis need to be spatially aggregated from 0. Because each 0. The typical approach to do so is weighting based on the inverse of the uncertainty:.
However, these formulations do not account for the spatial auto-correlation generated by the moving window 1 to 20 pixels may be common between two nearby estimates depending on the possible overlap of their respective 5 by 5 windows. This auto-correlation problem may be compounded further when only a clustered set of 0. To tackle this auto-correlation, we employ a more generic weighting approach. This information is summarized in a by matrix R a containing the fraction of overlap between every pair of windows. When the windows have no auto-correlations, both Eqs.
The aggregation procedure is applied to all data layers. Despite all efforts to characterize uncertainty and reach representative values, the results can still contain unrealistic values. A reason for this might be that uncertainties in the input data the remote sensing biophysical variables and the vegetation cover fraction maps are not explicitly taken into account.
As a final step to remove possible outliers, we remove all values for grid cells in which there are not at least 20 samples at 0. Lastly, we also remove values that are statistical outliers based on the distribution of the entire data set. All data layers are available with their associated uncertainty. Supplementary Fig. The local unmixing step can only be applied to those variables available at the 0.
The full surface energy balance is expressed as:. For the specific goals of this analysis we are interested in how the terms of this equation change according to a change in vegetation cover, i. We make the assumption that the changes in vegetation cover that are considered here are too small i. Although we derived estimates of changes in upwelling longwave flux satellite measurements at 0.
- Galiardo in G Major;
- Smart Fibres, Fabrics and Clothing: Fundamentals and Applications (Woodhead Publishing Limited series on fibers).
- Earth’s Ever-Changing Surface | Lesson Plan | PBS LearningMedia.
By re-writing and simplifying the equation above, the expression describing the change in the residual flux, composed of both sensible and ground heat fluxes, becomes:. Although a validation of the methodology against ground-based measurements of surface energy balance fluxes would be desirable, no adequate network of measurements currently exists.
Flux-towers would constitute the right measurements, but we would need a large and well distributed number of paired flux-tower sites with contrasting vegetation types yet similar climate, which currently do not exist. A comparison over a handful of sites 30 indicate that the results are in the right direction, even if the low number of pair-sites does not allow a robust and comprehensive verification of our data set. In the absence of validation, we propose a diagnostic to evaluate the robustness of latent heat flux MOD16A2 , arguably the most questionable input product, against an alternate data-driven product GLEAM v3.
We assume here that gains originate equally from all vegetation types suffering a loss. The assumption here is that the biophysical effects of vegetation cover change estimated from the — period remain valid for the — period. The resulting values are integrated in time to provide an estimate in exaJoules 10 18 Joules of the energy change caused by the given vegetation cover transitions. The surface energy balance can be decomposed to isolate the respective contribution of each component to the surface temperature resulting from the vegetation cover transition Such a method has been used to separate direct contributions e.
While only the direct effects can be separated in our data, it provides an easier way to interpret the energy effects caused by vegetation cover change from to To do so, we use Eqs. Calculating the derivative of Eq. The changes in downwelling radiation are neglected as before, leaving the changes in surface reflection, in latent heat and in the residual fluxes.
To connect with the global actual energy changes from to calculated before, the latter are transformed back into fluxes that can be used in Eq. To do so, the energy values in Joules are divided by the total changed area for each transition, and divided again by the number of seconds in a year, to obtain estimates of the annual radiative forcing at the surface that these changes have caused.
A number of assumptions were necessary to make our assessment which need to be taken into account when interpreting the results. To close the surface energy balance locally, we assume that the local cover change at 0. The robustness of this assumption relies on the fine scale of the analysis and on the typical lateral movement of air masses due to wind that ultimately advect air masses and clouds to different grid cells. We also assume vegetation cover is the only driver of changes in surface biophysics within the local moving window of 0.
For this purpose areas with strong elevation gradients are masked out to filter topographic effects, whereas any spatial gradient in general soil properties within the moving window is not considered. In our analysis of past changes, we consider that the biophysical signal of land cover change derived from observations acquired in — are representative for the entire — period.
This assumption holds if the background climate does not change substantially 50 , and could further be used to explore biophysical impacts in the near-future, but would require special attention to projecting them in a changing climate e. This assumption relies on the dominant role that climate variability has on climate trends on the decadal time scale. Finally, the biophysical effects considered here result from analyzing average conditions over 5 years — , while changes in more extreme years might be amplified. The numbers presented in this assessment are averaged both temporally and spatially, hiding part of the wealth of information generated in the underlying data.
Every record in space and time is also accompanied by an estimate of the uncertainty associated with the methodology that can serve to assess the relative quality of each value. Global Carbon Budget Earth Syst. Data 8 , — Myhre, G. Anthropogenic and natural radiative forcing. Canadell, J. Managing forests for climate change mitigation. Science , — Bonan, G. Forests and climate change: forcings, feedbacks, and the climate benefits of forests. Anderson, R. Biophysical considerations in forestry for climate protection.
Mahmood, R. Land cover changes and their biogeophysical effects on climate.
GEOG 272 Earth's Changing Surface (3 credits)
Lee, X. Observed increase in local cooling effect of deforestation at higher latitudes. Nature , — Davin, E. Climatic impact of global-scale deforestation: radiative versus nonradiative processes. Zhao, K. Biophysical forcings of land-use changes from potential forestry activities in North America.